Dual-targeted glutathione-glutamate functionalized Bismuth-Niosomes hybrid nanosystem for co-delivery of doxorubicin and Pi3K inhibitor into U87 glioblastoma cells

Abstract

The simultaneous use of multiple therapeutic methods as combinational therapy could be a promising strategy to overcome glioblastoma resistance. We introduced a hybrid Bismuth-Niosomes nanocarriers containing Pi3K inhibitor and doxorubicin drug and targeted by glutathione and glutamate ligands (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu). Finally, its anticancer effects were evaluated the in vitro on U87 glioblastoma cell line. The nanosystems were synthesized using a modified thin-film hydration technique and characterized using FTIR, UV-vis, DLS, TEM, FESEM, EDX, in vitro release, and hemolysis tests. The anticancer efficacy of nanosystems was evaluated through cellular uptake, cell viability, gene expression, cell-cycle, apoptosis, and scratch assays under both treatment conditions (without/with 2 Gy X-irradiation). Characterization results such as proper interaction among agent groups, nanometer size, negative charge, spherical morphology, good stability in biological medium, time/pH-dependent drug release, hemocompatibility, and high cellular internalization confirmed the correct synthesis of nanosystems. The treatment of U87 cells with BiNPs-NISM-Dox-Pi3Ki@BSA-GSH-Glu nanosystem under both conditions of X-ray exposure led to cell viability reduction, Vimentin, Cyclin D1, and Bcl-xL genes downregulation, cell cycle arrest (sub-G1 phase), apoptosis induction, and cell migration inhibition. The synergistic effects of final formulation of nanosystem along with X-irradiation exposure is statistically significant compared to No X-irradiation condition. The obtained results revealed that BiNPs-NISM-Dox-Pi3Ki@BSA-GSH-Glu nanosystem could be considered as a novel promising combinational therapy approach to overcome conventional limitations of current glioma cancer treatments through targeted drug delivery, signaling pathway inhibition, chemoradiotherapy. Further investigations including in vivo and clinical studies for assess reducing the side effects of drugs and optimum efficacy is required.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-54715-6.

Introduction

Glioblastoma multiforme (GBM) represents the most aggressive and lethal primary brain malignancy, characterized by rapid cellular proliferation, extensive invasiveness, and profound resistance to conventional therapeutic interventions. With an incidence rate of approximately 3–5 per 100,000 individuals annually and a median survival of merely 12–15 months post-diagnosis, GBM accounts for 14.5% of all central nervous system tumors and 48.6% of malignant brain cancers. The pathophysiology of glioblastoma is characterized by several hallmark features, including cellular heterogeneity, extensive angiogenesis, and the presence of an impermeable blood-brain barrier (BBB) that significantly restricts therapeutic drug penetration. The phosphoinositide 3-kinase (Pi3K)/Akt signaling pathway emerges as a critical oncogenic driver in over 80% of glioblastoma cases, promoting cellular survival, proliferation, and therapeutic resistance^1–3.

The clinical urgency of developing innovative therapeutic strategies for glioblastoma cannot be overstated, as current standard-of-care protocols involving surgical resection followed by radiotherapy and temozolomide chemotherapy provide only marginal improvements in patient outcomes. Novel approaches, such as tumor targeting, immunotherapy, and proton therapy, can give limited survival, but are mainly in the experimental stage, and their long-term efficacy is uncertain. In glioblastoma, the presence of a BBB and a microenvironment that can suppress the immune system can interfere with successful treatment^4,5. Doxorubicin (Dox) is a potent inhibitor of topoisomerase II, which induces DNA damage and apoptosis. However, its clinical use is limited by poor penetration of the BBB^6,7. These challenges highlight the critical need for novel approaches that can precisely target tumor cells and increase drug access^8,9. Combination therapy in cancer treatment refers to the simultaneous or sequential use of multiple therapeutic agents targeting distinct oncogenic pathways to enhance efficacy. It can also overcome resistance mechanisms to reduce the toxicity caused by high drug doses that occur with monotherapy. Using this approach can increase the effectiveness of treatment. This technique can increase treatment effectiveness. This technique takes advantage of synergistic interactions, which address tumor heterogeneity, compensatory survival signaling, and therapeutic escape in order to improve clinical outcomes^10,11.

In glioblastoma, the Pi3K/Akt/mTOR pathway is active in > 80% of cases, causing proliferation, survival, and resistance to treatment¹². Pi3K inhibitors (e.g., LY294002) block this pathway, suppressing pro-survival signals and sensitizing tumor cells to chemotherapy^13,14. Preclinical studies demonstrate that Pi3K inhibition downregulates anti-apoptotic proteins such as Bcl-2¹⁵ and enhances DNA damage by doxorubicin, a topoisomerase II inhibitor^16,17. Simultaneous delivery of these agents in a single nanocarrier ensures coordinated biodistribution, minimizing off-target toxicity while maximizing synergistic cytotoxicity.

The glutamate receptor, particularly α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, is abnormally activated in GBM, promoting tumor proliferation and invasion¹⁸. In this regard, glutamate-targeted nanoparticles can bind to AMPA receptors and facilitate receptor-mediated endocytosis. Also, the BBB can be a major obstacle to drug delivery. The urgent need for simultaneous BBB penetration and tumor-specific targeting justifies the rationale for using dual targeting by glutathione (GSH) and glutamate (Glu), which can increase specificity and intratumoral accumulation. Na-dependent glutathione transporter is mainly localized on the luminal membrane of brain, kidney, and small intestine endothelial cells. In contrast, its presence on endothelial cells of most other organs is comparatively limited^19,20.

Despite the recognized potential of combination therapeutic approaches, significant knowledge gaps remain regarding the best delivery strategies and targeting mechanisms for glioblastoma treatment. Current nanocarrier systems face significant challenges in effectively crossing the BBB while ensuring targeted drug release within the tumor microenvironment. The development of dual-targeted delivery systems that can simultaneously utilize multiple receptor-mediated transport mechanisms across the BBB is still largely unexplored. Additionally, combining radiosensitizing agents with chemotherapeutic drugs in hybrid nanocarrier systems presents unique formulation challenges that require innovative materials science solutions. The limited understanding of how GSH and Glu dual-targeting can improve selective glioblastoma cell uptake and BBB transport is a critical research gap that must be addressed to move precision nanomedicine forward²¹. This research introduces a groundbreaking hybrid nanosystem that addresses these limitations through the innovative co-delivery of Pi3Ki and doxorubicin using dual-targeted glutathione-glutamate functionalized Bismuth-Niosomes nanocarriers²². The primary objective is to develop a multifunctional therapeutic platform that combines the radiosensitizing properties of bismuth nanoparticles with the superior drug encapsulation capabilities of Niosomes (NISM), while leveraging dual ligand targeting for enhanced BBB penetration and tumor-specific delivery²³. Bismuth nanoparticles (Z = 83) are incorporated into the nanosystem to address therapeutic resistance by utilizing the radiosensitization potential of the high-Z element. This action provides a concentrated augmentation of energy delivery at the tumor location during radiotherapy²⁴. Preclinical models show that BiNPs increase radiotherapy efficacy in tumors while sparing healthy tissue^25,26.

This research represents a paradigm shift from conventional single-agent therapies toward precision combination nanomedicine, targeting both the Pi3K survival pathway and DNA integrity simultaneously. The synergistic integration of chemotherapy, pathway inhibition, and radiosensitization within a single delivery vehicle offers unprecedented potential for overcoming the therapeutic resistance mechanisms that have historically limited glioblastoma treatment efficacy. The glutathione-mediated BBB transport and glutamate receptor-targeted cellular uptake mechanisms provide a rational basis for achieving preferential tumor accumulation while minimizing systemic toxicity.

Materials and methods

Fetal bovine serum (FBS) (ES-020-B), 3-(4,5- dimethylthialzol-2-yl)−2,5-diphenyltetrazolium bromide (MTT) (57360- 69-7), Dulbecco’s modified Eagle’s medium, high glucose (DMEM) (D5796; Sigma). Cholesterol, Tween 80 (polyoxyethylene sorbitan mono-oleate), Span 80 (sorbitan mono-oleate), doxorubicin, and the Pi3K inhibitor (LY294002) were purchased from Sigma-Aldrich (MO, USA). Bovine Serum Albumin (BSA), Sodium hydroxide (NaOH), and Acetic acid, glutathione (GSH), glutamate (Glu), chloroform, and methanol were purchased from Merck (Darmstadt, Germany). U87 MG cell line (ATCC HTB-14) was obtained from the Stem Cell Research Center (Bon Yakhteh-Tehran, Iran). HEK-293 cell line (ATCC CRL-1573) was obtained from the Pasteur institute of Iran. Trypsin-EDTA (T3924) and Penicillin- streptomycin (P4333) were obtained from Sigma Co. (St Louis, MO, USA). Acetone, Dimethyl sulfoxide (DMSO), and Methanol were obtained from Tamadkala (Tamadkala Co., Iran). Phosphate-buffered saline (PBS) was prepared in the laboratory.

Preparation of bismuth nanoparticles (BiNPs)

BiNPs were synthesized via an environmentally benign protein-mediated reduction process adapted from established green synthesis methodologies. The synthetic approach exploited the dual functionality of bovine serum albumin (BSA) as both a reducing and stabilizing agent through redox interactions with bismuth (III) nitrate precursor. Initially, a 10% (w/v) BSA solution was prepared in 10 mL of phosphate-buffered saline (PBS, pH 7.4) under continuous stirring to ensure complete protein dissolution. Bismuth (III) nitrate pentahydrate (0.4% w/v) was subsequently added dropwise to the BSA solution under vigorous magnetic stirring (200 rpm) at room temperature to achieve homogeneous distribution and prevent localized precipitation. The reaction mixture pH was carefully adjusted to 8.5 using 1 N sodium hydroxide to optimize the reduction kinetics, followed by thermal incubation at 90 °C for 1 h with constant agitation (200 rpm) to facilitate nanoparticle nucleation and growth. The completion of BiNPs synthesis was visually confirmed by a characteristic color transition from colorless to deep black, indicating the formation of metallic bismuth nanostructures. For purification and separation of unreacted precursors and excess protein, the reaction mixture was subjected to preliminary centrifugation at 1000 rpm for 5 min to precipitate undissolved BSA aggregates, with subsequent collection of the nanoparticle-enriched supernatant. Further purification was achieved through extensive dialysis using a molecular weight cut-off membrane (MWCO 12 kDa) against ultrapure water for 24 h with periodic buffer exchange to remove residual salts, unreacted bismuth ions, and protein byproducts. The final product comprised a stable colloidal suspension of BSA-capped bismuth nanoparticles synthesized under biocompatible conditions²⁷.

Preparation of niosomes (NISM) and BiNPs-NISM nanocarrier (blank and encapsulated by therapeutic agents)

Niosomes were synthesized employing a modified thin-film hydration technique, a well-established methodology for vesicular drug delivery system fabrication. The lipid composition comprised 80 mg cholesterol and 600 mg of a carefully balanced surfactant blend containing Span 80 (sorbitan monooleate) and Tween 80 (polyoxyethylene sorbitan monooleate) in an equimolar ratio (1:1). This specific ratio was selected to achieve optimal hydrophilic-lipophilic balance (HLB) and ensure vesicular stability. These components were completely dissolved in 5 mL of chloroform within a 50 mL round-bottom flask containing glass beads to facilitate uniform film formation during subsequent rotary evaporation. The organic solvent was systematically removed using a rotary evaporator (Heidolph, Germany) operated at 60 °C under reduced pressure (474 mbar) for 20 min, resulting in the formation of a uniform thin lipid film on the flask walls. Complete elimination of residual chloroform traces was ensured through subsequent vacuum drying at 50 °C for 2 h. For the hydration process, an aqueous solution containing 0.02% (w/v) bovine serum albumin (BSA) was prepared in ultrapure water under constant magnetic stirring (150 rpm) for 15 min at ambient temperature to ensure complete protein dissolution. Subsequently, an appropriate volume of this BSA-containing distilled water was added dropwise to the thin lipid film under simultaneous stirring and sonication conditions to prepare NISM@BSA nanocarriers. Similarly, for the synthesis of BiNPs-NISM@BSA nanocarriers, an aqueous phase containing 0.02% (w/v) BSA and 0.02% (w/v) bismuth nanoparticles (BiNPs) was added dropwise to the thin lipid film under identical stirring and sonication conditions²⁸.

The hydration procedure was carried out using bath sonication (Branson CPX2800H, 35 kHz). The dispersion was kept in an ice bath to prevent overheating and maintain vesicle integrity, while the suspension was sonicated for 10 min at 50 W in pulse mode (5 s on/5 s off). NISM@BSA, BiNPs-NISM@BSA, and all drug-loaded formulations (BiNPs-NISM-Dox@BSA, BiNPs-NISM-Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA, as well as their GSH/Glu-functionalized derivatives) were prepared using the same method.

To synthesize therapeutic agent-encapsulated formulations (BiNPs-NISM-Dox@BSA or BiNPs-NISM-Pi3Ki@BSA), an aqueous phase containing 0.02% (w/v) BSA, 0.02% (w/v) BiNPs, and either 0.1% (v/v) doxorubicin (Dox) or 0.02% (v/v) Pi3K inhibitor (Pi3Ki) was maintained under constant magnetic stirring at room temperature for 2 h to ensure complete drug dissolution and equilibration. Subsequently, this drug-containing solution was added dropwise to the thin lipid film under simultaneous stirring and sonication conditions. The resulting suspension was incubated under identical conditions for an additional 2 h to ensure complete formulation maturation and drug encapsulation. For the preparation of dual-drug-loaded formulations (BiNPs-NISM-Dox/Pi3Ki@BSA), the aqueous phase contained 0.02% (w/v) BSA, 0.02% (w/v) BiNPs, 0.1% (v/v) Dox, and 0.02% (v/v) Pi3Ki. The same methodological approach was employed, wherein the multi-component aqueous phase was maintained under constant stirring for 2 h before dropwise addition to the lipid film, followed by synchronized stirring and sonication to promote vesicle formation and therapeutic agent entrapment. All formulations were stored at 4 °C under sterile conditions until further characterization and biological evaluation²⁹.

BiNPs-NISM@BSA nanocarrier functionalized with GSH/Glu

The glutathione-conjugated formulation (BiNPs-NISM@BSA-GSH) was prepared by incubating the BiNPs-NISM@BSA nanocarriers with 3.07 g (10 mmol) of GSH in the presence of 1% glutaraldehyde as a cross-linking agent. The reaction mixture’s pH was meticulously adjusted to 8.5 using a 1 M NaOH solution. The reaction was maintained under continuous stirring at room temperature for 24 h to ensure the successful binding of the GSH ligand to the nanocarriers. To incorporate the L-glutamate ligand and synthesize the dual-ligand formulation (BiNPs-NISM@BSA-GSH-Glu), the same synthetic procedure was followed, but the ligand mixture was adjusted. Specifically, the reaction utilized 1.54 g (5 mmol) of GSH and 0.74 g (5 mmol) of L-glutamate. The subsequent steps, including pH adjustment and incubation time, were identical to those used for the BiNPs-NISM@BSA-GSH preparation. Following synthesis, both formulations underwent extensive purification via dialysis. A molecular weight cut-off (MWCO) membrane of 12 kDa was used against ultrapure water for a total duration of 24 h, with periodic buffer exchange to remove all residual, unreacted reagents effectively³⁰.

Characterization of synthesized nanosystems

FT-IR and UV–vis spectroscopy analysis

To elucidate the chemical structure of the synthesized materials and investigate potential molecular interactions between the therapeutic agents and the nanocarriers, Fourier Transform Infrared (FT-IR) spectroscopy was employed. FT-IR spectra were acquired using a Bruker Tensor 27 spectrometer (Germany). The instrument operates within the mid-infrared range (4000–400 cm^− 1), a spectral region highly valuable for identifying functional groups and characteristic chemical bonding patterns.

The samples were meticulously prepared by dispersing the material within a matrix of potassium bromide (KBr). KBr was selected as the preferred matrix due to its chemical inertness and its transparency across the mid-infrared spectrum, which prevents interference with the distinctive absorption bands of the sample. The sample-KBr mixture was subsequently compacted into a transparent disc to ensure a uniform surface for infrared analysis. The resulting spectra were analyzed to confirm the presence of specific functional groups, determine bonding configurations, and identify potential interactions among the components: the NISM nanocarriers, BiNPs, Dox, Pi3Ki, GSH, Glu, and the BSA coating³¹. Furthermore, the optical characteristics of all prepared samples were evaluated using ultraviolet–visible (UV–vis) spectroscopy with a Shimadzu UV-160 Spectrophotometer (Japan)³².

Hydrodynamic size and zeta potential analysis

The hydrodynamic average size (D_h), polydispersity index (PDI), and zeta potential (ζ) of the nanoparticles integrated within the drug delivery system (DDS) were rigorously assessed using Dynamic Light Scattering (DLS) and electrophoretic light scattering techniques. These parameters are fundamental for evaluating the physicochemical stability and uniformity of the nanocarrier suspensions, which critically influence their biological interactions and drug-carrying efficacy. The DLS and zeta potential analyses were conducted using a Malvern Nano ZS device (Malvern Instruments, Worcestershire, UK), a high-precision instrument widely recognized for nanoparticle characterization. For sample preparation, a 0.5 mL aliquot of the nanocarrier suspension was precisely diluted with 2 mL of deionized water. This dilution step was performed directly within a sterile Malvern sample vial to mitigate potential contamination that could skew the measurements. Upon completion of the dilution, the DLS analysis commenced, providing detailed insights into the suspended nanoparticles’ behavior.

The instrument determined the average hydrodynamic diameter (D_h) by measuring the Brownian motion of the particles. This D_h represents the size of the nanoparticle core plus the layer of solvent molecules dynamically associated with its surface. Concurrently, the PDI was calculated. As a dimensionless parameter, the PDI quantifies the breadth of the size distribution within the nanoparticle sample. A low PDI is indicative of a narrow, more uniform size distribution, which is generally desirable for therapeutic applications as it helps ensure reliable drug delivery and predictable biodistribution. Finally, the zeta potential (ζ)—a measure of the nanoparticles’ surface charge—was determined. This parameter is particularly crucial as it dictates particle stability in suspension. A high magnitude of ζ (either strongly positive or strongly negative) generates significant electrostatic repulsion between the particles, which effectively counteracts attractive forces and minimizes the propensity for aggregation and subsequent destabilization³³.

Morphological and dimensional analysis

The morphological characteristics and particle dimensions of the synthesized bismuth nanoparticles and NISM nanocarriers were rigorously analyzed. Field Emission Scanning Electron Microscopy (FESEM) was performed using a TESCAN MIRA3 (Brno, Czech Republic) to examine the surface morphology and particle shape meticulously. Prior to imaging, specimens were coated with a thin layer of gold to enhance conductivity and image resolution. Further investigation into the nanoparticle dimensions and ultrastructure was conducted using Transmission Electron Microscopy (TEM). A Philips EM208S electron microscope, operating at 100 kV, was employed for this analysis³⁴.

Long-term nanosystem colloidal stability assessment

The colloidal stability of the hybrid nanoparticles was systematically evaluated to ensure their suitability for biomedical applications requiring an extended shelf-life. This assessment involved monitoring key physicochemical properties using DLS over a period of approximately two months under refrigerated storage conditions (4 ^ͦC). The critical parameters tracked were the hydrodynamic diameter (D_h), zeta potential (ζ), and PDI. Formulations were quantitatively classified as stable if they maintained a size variation of less than 15% in D_h and an associated PDI value of less than 0.3 throughout the testing period. Formulations considered acceptable provided they satisfied the following predefined criteria: (i) PDI < 0.3 during the observation period and (ii) relative variation in Dh < 15% compared with the baseline value. These parameters were chosen to prevent aggregation-driven particle population broadening and to ensure the preservation of a confined size distribution. To make sure uniform redispersion, all batches were gently vortexed before measurement and kept at 4 °C under sterile conditions³⁵.

Encapsulation efficiency determination of Dox and Pi3Ki in the nanosystem

The encapsulation efficiency (EE%) of Dox and Pi3Ki within the synthesized nanosystem was quantified using UV-vis spectrophotometry, a procedure consistent with previously validated analytical techniques. This methodology relies on the indirect determination of the encapsulated drug mass by accurately measuring the amount of non-encapsulated (free) drug present in the suspension supernatant. Initially, to ensure accurate quantification, standard calibration curves were constructed for both therapeutic agents. Doxorubicin standards were prepared at concentrations spanning from 3.125 to 200 µg/mL, while Pi3Ki standards covered a lower concentration range of 0.5 to 16 ng/mL.

These curves established the necessary linear relationship between drug concentration and absorbance. Following the drug loading stage, the nanosystem suspension underwent a crucial separation step to isolate the free drug. This was performed using a centrifugation-ultrafiltration technique. The suspension was centrifuged at 4000 rpm through a filter unit equipped with a 12 kDa molecular weight cut-off (MWCO) membrane. This specific pore size was selected to ensure that the larger, drug-loaded nanosystems were retained by the filter while only the small, non-encapsulated Dox and Pi3Ki molecules passed through into the collected supernatant. The absorbance of the resulting supernatant, containing the free drug, was then measured using the UV-vis spectrophotometer at the respective maximum absorption wavelengths (λ_max). Free Dox concentration was quantified by measuring absorbance at λ_max = 492 nm, and Pi3Ki was quantified at λ_max = 294 nm. The measured absorbance values were subsequently referenced against the previously established standard calibration curves to determine the absolute mass of non-encapsulated drug accurately.

The EE% for each drug was then calculated based on the difference between the initial total drug mass added during formulation and the mass of the non-encapsulated drug determined in the supernatant³⁶. This relationship is formally expressed by Eq. 1:

1

Where W_t is the total amount of drug in the nanovesicle suspension, and W_i is the total quantity of drug added initially during preparation.

Conjugation efficiency determination of GSH and Glu ligands to hybrid nanosystems

The conjugation efficiency of the targeting ligands, GSH and Glu, to the BiNPs-NISM@BSA nanocarriers was quantitatively determined using UV-vis spectrophotometry. This analytical approach focused on measuring the concentration of the unreacted ligands remaining in the solution after the functionalization process. Before sample analysis, standard calibration curves were constructed for both GSH and Glu to establish the linear correlation between concentration and absorbance, enabling the accurate quantification of unknown samples. GSH standards were prepared over a concentration range of 20 to 100 µM, while Glu standards spanned from 20 to 200 µg/mL. Following the conjugation reaction, the functionalized nanosystems were separated from the surrounding solution containing the unreacted ligands. The resulting supernatant, or unbound solution, was then collected and subjected to spectrophotometric analysis. The absorbance of the unbound solution was precisely measured at the respective maximum absorption wavelengths (λmax): GSH at 500 nm and Glu at 216 nm. By referencing these absorbance readings against the standard calibration curves, the precise mass of unbound GSH and Glu was accurately determined. The conjugation efficiency (CE%) for each targeting compound (GSH or Glu) was calculated based on the difference between the initial total mass of the compound introduced into the reaction and the mass of the unreacted compound quantified in the supernatant (Eq. 1)³⁷.

In vitro drug release kinetics of Dox and Pi3Ki

The in vitro release profiles of Dox and the Pi3Ki from the nanosystem were systematically investigated using the dialysis membrane diffusion technique across different pH conditions, simulating distinct physiological microenvironments³⁸.

In vitro drug release assay

A predetermined quantity of the nanosystems encapsulating both Dox and Pi3Ki was loaded into a dialysis bag. The drug-loaded nanosystems were initially suspended in 2 mL of phosphate buffer saline (PBS) containing 10% ethanol. This dialysis bag was subsequently immersed in an external release medium consisting of 30 mL of PBS (10% ethanol). The release study was conducted at a physiological temperature of 37 ͦ C under continuous agitation (100 rpm) using a shaking incubator. To evaluate the pH-responsive release behavior, two distinct pH values were maintained for the release medium: pH 7.4 (simulating normal blood pH and tumor extracellular pH) and pH 5.8 (simulating the acidic environment within endosomes and lysosomes of cancer cells). At predetermined time intervals, a 2 mL aliquot of the external release medium was withdrawn for quantification. The concentrations of Dox and Pi3Ki were measured using a UV-vis spectrophotometer at their respective maximum absorption wavelengths (λ_max): 492 nm for Dox and 294 nm for Pi3Ki.

Crucially, following each sampling, the 2 mL volume was replaced with fresh release medium to ensure sink conditions were maintained throughout the experiment, thus preserving a constant volume for the drug release test.

Drug release kinetic analysis

To elucidate the underlying mechanisms governing the release of Dox from the nanosystems and predict their long-term behavior, the experimental in vitro drug release data were fitted to various drug release kinetic models (as summarized in Table S1). The goodness-of-fit for each model was rigorously assessed by calculating the mean squared error (MSE) and Akaike’s Information Criterion (AIC).

The kinetic model that yielded the lowest MSE and AIC values was designated as the best-fit model, providing a quantitative framework for characterizing the Dox release mechanism³⁹.

Hemocompatibility assay

The hemocompatibility of the treatments was evaluated by quantifying their hemolytic activity against human red blood cells (HRBCs), following a previously validated protocol⁴⁰. All procedures involving human participants were performed in accordance with the ethical standards of the institutional and national research committees, the 1964 Declaration of Helsinki and its later amendments, and comparable ethical standards. The hemocompatibility study protocol, including the collection and use of human red blood cells, was reviewed and approved by the Ethics Committee of Zanjan University of Medical Sciences, Zanjan, Iran (Ethical Code: IR.ZUMS.BLC.1402.027). Written informed consent was obtained from the blood donor (principal investigator) before sample collection, including consent for the use of the samples for research and publication of anonymized data. The collected blood samples were subsequently processed to isolate the red blood cells for following evaluations. All procedures were executed according with institutional protocols and ethical standards. This assay determines the potential for the nanosystems to induce red blood cell lysis, which is a critical safety parameter for intravenous administration. The experimental protocol began by preparing a 0.5mL total volume mixture, which consisted of 0.25 mL of a RBC suspension combined with 0.25 mL of the corresponding testing solution. The assay utilized two primary controls: Triton X-10 served as the Positive Control to establish a baseline for 100% hemolysis, while Phosphate-Buffered Saline (PBS) served as the Negative Control to determine the 0% hemolysis baseline.

The two nanosystem formulations the unloaded nanosystem (BiNPs-NISM@BSA) and the dual-drug-loaded, targeted nanosystem (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu)—were assessed across a concentration gradient of 5, 15, 30, 45, and 90 µg/mL. Following sample preparation, all mixtures were incubated at a constant temperature of 37 ͦ C for a period of 4 h under continuous shaking to simulate the physiological environment of blood circulation. After this incubation phase, the samples were subjected to high-speed centrifugation at 2800 rpm for 15 min to effectively separate intact cells and nanoparticles from the liquid phase containing any potentially released hemoglobin. The percentage of hemolysis was determined using the following equation (Eq. 2), which normalizes the absorbance of the released hemoglobin in the sample against the maximal release observed in the positive control, Eq. 2:

2

In this equation, A _{treated sample}, A _{negative control,} and A _{positive control} are representative of the mean absorbance of the sample, negative control, and positive control, respectively.

Cell culture procedure

U87 MG cell line (ATCC HTB-14) and HEK-293 cell line (ATCC CRL-1573) were obtained from the Stem Cell Research Center (Bon Yakhteh-Tehran, Iran) and the Pasteur institute of Iran, respectively. The cells were cultured in DMEM, high glucose, supplemented with 10% FBS, 100 µg/mL penicillin, and 100 µg/mL streptomycin. The cells were maintained in 25 cm² cell culture flasks under standard conditions at 37 °C in a humidified atmosphere containing 5% CO₂. The culture medium was replaced every 48–72 h, and cells were passaged at 80–90% confluency using 0.25% trypsin-EDTA solution.

Cellular uptake assay

The U87 (high glutamate receptor expression) and HEK-293 (low glutamate receptor expression) cell lines were used which seeded in 12-well plates (at a density of 3.6 × 10⁴ cells per well (with 500 µL of complete medium and allowed to grow overnight under standard conditions (37 °C, 5% CO₂). U87 cells were divided into untreated (Control) and treatment groups, as separately, exposed to 500 µL of medium containing BiNPs-NISM@BSA (10 µg/mL), BiNPs-NISM-Dox/Pi3Ki@BSA (10 µg/mL), or BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu (1, 2.5, 5, 10 µg/mL based on Dox and Pi3Ki), while HEK-293 cells only were treated with 10 µg/mL of BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu. After 24 h, cells were washed with PBS, trypsinized, centrifuged (1200 rpm, 5 min), resuspended in 500 µL PBS, and analyzed via flow cytometry (BD Biosciences, San Jose, CA) with data processed using FlowJo v7 software (Tree Star, Ashland, OR)⁴¹.

Cell viability (MTT) assay

The viability effects of synthesized nanosystems were evaluated against U87 glioblastoma cells using the MTT assay. The cells were seeded in 96-well plates at a density of 1.2 × 10⁴ cells/well in high-glucose DMEM supplemented with 10% FBS and allowed to grow overnight under standard conditions (37 °C, 5% CO₂). The cells were treated with blank nanocarrier (BiNPs-NISM@BSA), Free Dox drug, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu in 2.5, 5 and 10 µg/mL concentration. Following a 5-hour incubation period, the culture medium was carefully replaced with 500 µL of fresh complete medium in each well. Subsequently, the cultured cells were exposed to fractionated X-ray radiation at a dose of 2 Gy. After 24 h of treatment, the medium was carefully removed from all experimental groups, and 20 µL of MTT solution (5 mg/mL) was added to each well for 3–4 hours.

After removing the MTT solution, the cell pellets were dissolved in 100 µL of DMSO per well. Absorbance measurements were subsequently recorded at 570/630 nm using an ELISA reader. All steps as mentioned above were also performed without X-ray exposure²⁹.

Real‑time PCR assay

For gene expression analysis, U87 cells were cultured in T-25 flasks (2 × 10⁶ cells/flask) and divided into four treatment groups: (1) untreated control, (2) free Dox drug, (3) niosome-bismuth hybrid nanoparticles with drugs (BiNPs-NISM-Dox/Pi3Ki@BSA), and (4) dual-targeted nanoparticles (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu). Following 24 h incubation (37 °C, 5% CO₂), total RNA was extracted using the RNX-Plus according to the manufacturer’s instructions, and RNA concentrations were measured via Nanodrop (A260/A280 ratio > 1.8)⁴². Complementary DNA (cDNA) was synthesized using the RevertAid First Strand cDNA Synthesis Kit. Quantitative real-time PCR (RT-qPCR) was conducted using the RealQ Plus 2x Master Mix Green according to the manufacturer’s protocol. This experiment was carried out on an Applied Biosystems™ Real-Time PCR Instrument (Thermo Fisher Scientific, USA). The gene expression levels of Vimentin, Cyclin D and BCL-xL were measured and GAPDH have been employed as the housekeeping reference gene. The primer sequences for the target genes are presented in Table S2.

Cell cycle assay

To assess cell cycle arrest after treatment with synthesized nanosystems, U87 glioblastoma cells were seeded in 12-well plates (6 × 10⁴ cells/well) and incubated overnight under standard conditions (37 °C, 5% CO₂). The cells were treated with 5 µg/mL of either BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu.

Then, 5 hours’ post-treatment, the medium from all cell groups were removed, and 500 µL of fresh complete medium was added to each well. The cells were then exposed to a 2 Gy fractionated X-ray. After 24 h total incubation, cells were harvested using 0.05% trypsin-EDTA, fixed in 70% ice-cold ethanol, and stored at −20 °C for two h. Before analysis, cells were washed with PBS and stained with 1 mL PI (propidium iodide) Master Mix solution (40 µL PI, 10 µL RNase, 950 µL PBS) for 30 min at 37 °C. Flow cytometry was performed using a FACSCalibur system (BD Biosciences, San Jose, CA), and the acquired data were processed with FlowJo v.7 (Tree Star, Ashland, OR). All steps as mentioned above were also performed without X-ray exposure⁴³.

Apoptosis assay

To assess the apoptosis induction after treatment with synthesized nanosystems, U87 glioblastoma cells were seeded in 12-well plates at a density of 6 × 10⁴ cells/well and allowed to grow overnight under standard conditions (37 °C, 5% CO₂). The cells were treated with 5 µg/mL of either BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu for 5 h. Untreated cells were regarded as the control. To assess the effects of X-ray irradiation on apoptotic rate, the culture medium was replaced with 500 µL of fresh growth medium in all wells. The culture plate was then subjected to 2 Gy radiation using a clinical linear accelerator, followed by a 24-hour incubation period under standard conditions. Then, cells were harvested using trypsin-EDTA, washed twice with cold PBS, and dual-stained with Annexin V-FITC (5 µL) and propidium iodide (PI, 2 µg/mL) for 15 min at room temperature in the dark. Flow cytometry was performed using a FACSCalibur system (BD Biosciences, San Jose, CA), and the acquired data were processed with FlowJo v.7 (Tree Star, Ashland, OR). All steps as mentioned above were also performed without X-ray exposure^28,44.

Scratch assay

The anti-migratory effects of synthesized nanosystems, on U87 glioblastoma cells were evaluated under both standard and irradiated conditions. The cells were seeded in duplicate 24-well plates at 3.6 × 10⁴ cells/well and grown to 85% confluence in complete medium. A uniform linear scratch was created using a sterile 10µL pipette tip, followed by three PBS washes to remove dislodged cells. The cells were then treated with a 5 µg/mL concentration of BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu. Then, 5 h after treatment, the cells were irradiated with a 2 Gy dose of fractionated X-rays. Finally, microscopic images of the scratched area were captured at 0 to 48 h and analyzed using Image J (v1.52; NIH) software²⁸. The migration inhibition rate (%) was calculated using the following formula (Eq. 3). All steps as mentioned above were also performed without X-ray exposure.

3

At=0 h is the area of the wound measured h hours after the scratch is performed.

At = Δh is the area of the wound measured immediately after scratching.

Statistical analysis

All quantitative data were analyzed using GraphPad Prism 9.0, and the data are presented as mean ± SD of at least three independent biological/technical replicates. For more than two groups, one-way ANOVA with Tukey’s post hoc test was applied, and for time- or dose-dependent research, two-way ANOVA with Sidak’s multiple comparisons test. Statistical significance levels were as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Results and discussions

The motivation for developing this new combination therapy approach involving dual-targeted glutathione-glutamate functionalized BiNPs-NISM@BSA nanosystems containing doxorubicin and Pi3K inhibitor, along with X-irradiation against U87 glioblastoma cells rests on compelling clinical-therapeutic rationales. The total ligand input (10 mmol) used for the single-ligand GSH system survived in the dual-ligand formulation, which was separated equally between 5 mmol GSH and 5 mmol L-glutamate (1:1 molar ratio). In order to preserve the optimal overall ligand loading, colloidal stability, and zeta potential established in the GSH-only formulation, this ratio was selected to provide a balanced surface density of both ligands, facilitating the nanosystem to efficiently engage glutamate receptors overexpressed on glioblastoma cells and glutathione transporters at the blood–brain barrier.

GBM, the most aggressive and deadly primary brain tumor, leaves patients with a woeful median survival of only 12–15 months’ post-diagnosis despite aggressive multimodal treatment options thereof comprising surgical resection, radiation therapy, and conventional chemotherapy. This dreadful prognosis truly warrants novel therapeutic strategies in the very least to counter the intrinsic limitations of the current treatment modalities⁴⁵. The therapeutic rationale encompasses critical killing considerations. Niosomes constitute a sophisticated class of lipid-based nano-vesicular drug delivery systems that have shown high promise in the encapsulation and delivery of classical chemotherapeutics, especially doxorubicin, while overcoming biological barriers to improve the therapeutic efficacy. The incorporation of the Pi3K inhibitor (LY294002) into this nanosystem is aimed at sensitizing glioblastoma cells to the chemo-agents by targeting the Pi3K/Akt survival pathway that is dysregulated in more than 80% of glioblastoma cases and which contributes significantly to therapeutic resistance mechanisms.

Chemotherapy in combination with radiotherapy has found utility and been better established as a clinically efficacious scheme maximizing synergism. The BiNPs included in the hybrid formulation serve their purpose in radiosensitization thanks to the high atomic number (Z = 83), which results in preferential dose deposition to the tumor tissues, sparing the healthy surrounding brain parenchyma. The radiosensitizer amplifies the therapeutic index of conventional radiotherapy protocols.

An important novelty in this nanosystem is the dual-ligand targeting via glutathione and glutamate functionalization. Treatment of glioblastoma represents a major challenge due in significant part to crossing the BBB, which restricts the uptake of therapeutic agents into the brain tissue. Transport facilitated by glutathione exploits the sodium-dependent glutathione transporter that is highly localized on the luminal membrane of the brain endothelial cells and helps the penetration of these agents into the brain through the BBB. At the same time, targeting of glutamate receptors banks on the overexpression of AMPA receptors by glioblastoma cell lines, which under abnormal conditions in GBM promote proliferation and invasion. This dual-target stimulator, therefore, provides both enhanced BBB translocation and tumor cell accumulation to maximize therapeutic specificity while minimizing systemic toxicity. The amalgamation of these various therapeutic regimens into one nanosystem brings about a paradigm change to precision combination nanomedicine targeting the Pi3K survival pathway, inducing DNA damage by doxorubicin, and enhancing radiosensitivity through bismuth incorporation. This approach addresses glioblastoma resistance mechanisms that are multifactorially implicated, providing a truly revolutionary prospect to improve treatment for one of the most persistent malignancies.

FT-IR spectroscopy analysis and molecular interaction assessment

Fourier Transform Infrared spectroscopy served as the most crucial analytical technique to confirm the successful synthesis of the novel dual-targeted BiNPs-NISM@BSA hybrid nanosystems, along with elucidation of the intermolecular interactions existing between the constituent components. Spectroscopic analysis rendered conclusive evidence for the introduction of BiNPs, Pi3K inhibitor, Dox, GSH, and Glu in the nanocarrier matrix while also deciphering the nature of intermolecular interactions stabilizing the multi-component therapeutic system. The FT-IR scans showed the presence of characteristic absorption bands, thereby confirming the successful incorporation of BiNPs into the NISM matrix beyond any doubt. The spectrum showed the fundamental vibrational modes at 3400, 2385, 1621, 1371, and 825 cm^-1, corresponding to O-H stretching, C-H stretching, O-H bending, C-C stretching, and C-O stretching vibrations, respectively. More importantly, the diagnostic absorption bands observed in the 600–550 cm^-1 region provided definitive evidence for the Bi-O-Bi bridging modes of crystalline bismuth nanoparticles, thus confirming their successful integration within the hybrid nanosystem. The appearance of peaks at 450 to 420 cm^-1 is associated with Bi-S stretching vibrations, which strongly support the formation of bismuth-thiol interactions when capped with BSA, which was previously reported in these systems (Fig. 1A)⁴⁶. The spectroscopic analysis of Pi3Ki encapsulation revealed complex molecular interactions that distinguish this formulation from traditional single-drug delivery systems. Characteristic absorption bands of free Pi3Ki were observed at approximately 3030 cm^-1 (aromatic C-H stretching), 1705–1690 cm^-1 (carbonyl C = O stretch), 1600–1580 cm^-1 (aromatic C = C stretches), 1500–1480 cm^-1 (C-N-C stretching), and 1250–1230 cm^-1, as well as 1110–1100 cm^-1 (C-O-C stretching) and 830–800 cm^-1 (fingerprint).

Fig. 1.

FT-IR spectra of synthesized nanosystems. (A) BiNPs, BiNPs-NISM@BSA, and GSH spectra, showing the successful formation of BiNPs-NISM@BSA and the interaction with GSH. (B) BiNPs-NISM-Pi3Ki@BSA-GSH, BiNPs-NISM@BSA, and Pi3Ki spectra, indicating successful encapsulation of Pi3Ki onto the nanosystem. (C) BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, BiNPs-NISM-Pi3Ki@BSA-GSH, and free Dox spectra, confirming co-encapsulation of Dox and Pi3Ki and comparing with free Dox.

Notably, the spectrum for drug-loaded nanocarriers showed an absorption band at 2950 cm^-1, attributed to C-H stretching, which was present in both free and encapsulated Pi3Ki. This indicates the structural integrity was maintained during encapsulation. The strongest spectroscopic evidence of successful encapsulation was observed in the 1400–1300 cm^-1 region, where peaks associated with aromatic C = C and C-N stretching appeared with significantly reduced intensity in the nanocarrier spectrum. This reduction, along with slight peak shifts of the C = O and aromatic C = C stretches, provides strong proof of hydrophobic interactions between Pi3Ki and the NISM bilayer. This offers new insight into how Pi3K inhibitors are encapsulated within BiNPs-NISM@BSA (Fig. 1B)⁴⁷.

FT-IR spectroscopic analysis of Dox incorporation provided comprehensive evidence of successful drug encapsulation and the molecular mechanisms governing drug-carrier interactions. Characteristic peaks of the Dox spectrum were observed at 3441 cm^-1 (O-H and N-H functional groups), 1634 cm^-1 (C = O functional group), 1561 cm^-1 (N-H functional group), 1417 cm^-1 (O-H functional group), and 1113 cm^-1 (C-O functional group), thereby validating literature reports⁴⁸. Characteristically, Dox encapsulation within the nanocarriers was altered substantially, spectroscopically speaking, giving operative evidence toward a mechanistic understanding of drug-carrier interactions. The reduction in band intensity and slight peak shifting for encapsulated Dox confirmed hydrogen bonding and electrostatic interactions formed between carbonyl and hydroxyl groups of Dox with hydrophilic domains of the NISM bilayer. These spectroscopic characteristics of drug encapsulation - peak broadening and attenuation -are indicators of restricted vibrational freedom following entrapment, constituting strong evidence for successful Dox incorporation into the BiNPs-NISM@BSA nanosystem (Figs. 1C and 2A)⁴⁹.

Fig. 2.

FT-IR spectra of synthesized nanosystems. (A) Comparison of free Dox, BiNPs-NISM@BSA, and BiNPs-NISM-Dox@BSA-GSH, showing successful encapsulation of Dox into the nanosystem. (B) BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, GSH, and BiNPs-NISM-Dox/Pi3Ki@BSA spectra, indicating the interaction with GSH and stability of the co-encapsulated system. (C) BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, Glu, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu spectra, demonstrating the effect of glutamate treatment on the nanosystem.

The FT-IR spectrum of GSH revealed characteristic vibrational modes that provided detailed molecular information about the targeting ligand incorporation. The broad absorption band at 3360–3270 cm^-1 was definitively assigned to N-H stretching vibrations, indicative of primary amine (-NH₂) and secondary amide (-NH-) functionalities. The diagnostic but weak peak at 2520 cm^-1 confirmed the presence of the thiol (-SH) group, while aliphatic C-H stretching vibrations appeared at 2950–2850 cm^-1.

Peptide backbone characteristics were highlighted by amide I-band 1690–1720 cm^-1 carbonyl C = O stretching and amide II band 1540–1530 cm^-1 N-H bending coupled with C-N stretching. More features included symmetric secondarily carboxylate (COO^-) stretching and CH^- bending at 1400–1380 cm^-1, amide III vibration within the 1220–1170 cm^{- 1} wavelength region, and hydroxyl/carboxyl group C-O stretching modes ranging from 1060 to 1040 cm^-150. Most significantly, the 970–900 cm^{- 1} region exhibited C-S stretching, serving as a definitive marker of the cysteine thiol moiety and confirming successful GSH conjugation to the nanosystem surface (Fig. 2B)⁵¹. The spectra indicated successful incorporation of Glu by reflecting unique patterns of absorptions characterizing effective dual conceptualization targeting. The Glu spectrum displayed some distinct peaks at 3490, 2850, 1690, and 1476 cm^-1 that were ascribed to the O-H stretching vibration, C-H stretching vibration, stretching of carbonyl, and N-H stretching of the amino group, respectively. Retention of these characteristic peaks in the end formulation was conclusive evidence for successful peptide conjugation; the structure of the targeting ligand was preserved (Fig. 2C)⁵².

UV-vis spectroscopy: molecular interaction analysis and synthesis validation

Ultraviolet-visible spectroscopy was rigorously employed as a rich analytical tool for the study of intermolecular interactions between the constituent components, as well as the validation of the successive synthesis of this new multi-targeted hybrid nanosystem. This spectroscopic approach provided completely convincing evidence for the successful integration of BiNPs, therapeutic agents, and targeting ligands while highlighting the underlying molecular interactions that impart stability and alter functionality upon the system.

The UV-visible characterization of bismuth nanoparticles showed an absorption peak at 242 nm, which is normal for the surface plasmon resonance (SPR) behavior of metallic bismuth nanostructures studied in the literature (Fig. 3A)⁴⁶. This diagnostic absorption band was still observed; however, the highly apparent red shift assigned spectroscopic evidence to the successful bismuth integration within the hybrid matrix. The observed red shift would also imply the formation of the bismuth-thiol coordination complex through BSA, thus demonstrating a successful green synthetic route and the hybridization of the BiNPs with NISM@BSA nanocarriers. This blue shift phenomenon is attributable to changes in the local dielectric environment surrounding the bismuth nanoparticles following their binding within the BSA-stabilized NISM nanocarriers. Such a shift is consistent with fundamental principles of nanoparticle surface modification, whereby protein adsorption alters the electronic properties and plasmonic resonance of the nanomaterial. Notably, previous studies have shown that bismuth nanoparticles exhibit size-dependent UV-vis absorption profiles, with distinct characteristics for particles below 10 nm^53,54.

Fig. 3.

UV–vis absorption spectra of synthesized nanosystems. (A) BiNPs, NISM@BSA, and BiNPs-NISM@BSA spectra, showing the successful formation of the nanosystem. (B) Pi3Ki, BiNPs-NISM@BSA, and BiNPs-NISM-Pi3Ki@BSA spectra, indicating successful encapsulation of Pi3Ki into the nanosystem. (C) Free Dox, BiNPs-NISM-Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA spectra, demonstrating co-encapsulation of Dox and Pi3Ki and their optical properties.

Our findings advance this understanding by demonstrating that surface coordination with BSA—particularly through its disulfide-rich domains—induces a blue shift in the absorption spectrum, thereby modifying the optical behavior of bismuth nanoparticles in these biological nanocomposites. Spectroscopic characterization of the encapsulation of the Pi3K inhibitor has provided a deeper understanding of drug-carrier interactions, particularly for loading efficiency assessments. Free Pi3K inhibitor suppressed a strong absorption maximum at 294 nm, which corresponds to electronic transitions in its morpholine-chromone molecular architecture (Fig. 3B). The encapsulated formulation only presented slightly less pronounced absorption compared to the drug-free nanosystem, lending quantitative credence to the successful loading of the inhibitor into the nanocarrier matrix. The intensity diminishment is believed to be a consequence of restricted motions and altered electronic transitions upon encapsulation into the lipid bilayer ambiance. In other words, the spectroscopic modifications observed are a thumbprint of the commonly accepted mechanism of drug hydrophobic encapsulation, whereby the inhibitors become trapped in the hydrophobic conduit of the NISM bilayer, affording additional optical peculiarities. This is, therefore, the first systematic UV-vis characterization of Pi3K inhibitor encapsulation within BiNPs-NISM@BSA nanosystems, elucidating the molecular rationale behind drug-carrier interactions within a precision combination nanomedicine platform.

The UV-vis spectroscopic evaluation of Dox encapsulation revealed characteristic absorption features that offer mechanistic insight into how the drug loads and is retained within the nanosystem. Free Dox showed a distinct absorption peak at 492 nm, caused by electronic transitions in the anthraquinone chromophore system (Fig. 3C).

Dox carries the fundamental electronic signature of the entire anthracycline class of chemotherapeutic agents and is essential for its DNA intercalation ability. The presence of the characteristic peak in the Dox-encapsulated nanosystem confirmed successful drug incorporation while maintaining the structural integrity of the active chromophore. The preservation of spectroscopic properties indicates that, during encapsulation, the defect in the conjugated π-electron system remains intact, which is critical for Dox’s anticancer function.

Therefore, spectroscopic validation supports that the hybrid nanosystem retains the inherent pharmacological properties of the encapsulated therapeutic agents—an essential consideration in developing combination nanomedicine⁵⁶.

The UV-vis spectroscopic analysis strongly indicates the successful conjugation of GSH and Glu as well as the dual targeting capabilities of hybrid nanosystems (Fig. 4A). The native, reduced form of GSH, a tripeptide composed of glutamate, cysteine, and glycine, exhibits no significant absorption peak in the standard UV-Vis range above 220 nm because its major functional groups—primarily the peptide bonds and the thiol (SH) group—only contain electronic transitions that occur at much shorter wavelengths, typically below 220 nm⁵⁷. The covalent conjugation of the GSH ligand to the nanocarrier surface is definitively confirmed by an observable bathochromic shift (red-shift) in the UV-vis spectrum of the resulting structure (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH). This distinct shift serves as a potent spectroscopic fingerprint for the successful chemical modification of the nanosystem.

Fig. 4.

UV–vis absorption spectra of synthesized nanosystems. (A) GSH, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH spectra, showing the interaction of the nanosystem with GSH. (B) Pi3Ki, BiNPs-NISM-Dox@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH spectra, indicating successful co-encapsulation of Pi3Ki and Dox. (C) BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, Glu, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu spectra, demonstrating the effect of glutamate on the nanosystem’s absorption properties.

The spectral alteration is attributed to a change in the electronic environment of the participating chromophores and the newly formed chemical bonds. Specifically, the conjugation process likely results in increased electronic delocalization within the GSH-nanocarrier complex, which effectively lowers the energy required for electronic transitions (). These molecular changes provide quantitative and qualitative evidence for the effective chemical binding of the targeting GSH ligand onto the BiNPs-NISM@BSA platform (Fig. 4B). As shown in Fig. 4C, the Glu-targeting ligand exhibited a clear absorption peak at around 350 nm, which corresponds to n→π* electronic transitions in amino acids⁵⁸. The preservation of such features in BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu further demonstrates that Glu was effectively incorporated, as evidenced by spectral data, in what appears to be a dual cancer-targeted carrier.

This platform combines glutathione and Glu targeting in a single nanocarrier. Significantly, upon Glu conjugation and dual targeting, shifts and changes in absorbance intensity are observed in peaks corresponding to the nanocarrier. Such differences likely point to newly emerging intermolecular interactions between the ligand and therapeutic agents. These spectral changes also indicate that surface functionalization using targeting ligands will not only offer biological recognition specificities but will also modify the associated local physicochemical environment, possibly affecting drug release profiles and uptake efficiency.

Hydrodynamic diameter analysis and sequential modification characterization

According to hydrodynamic size observations, the sizes were estimated at 161.5 ± 7.3 nm for BiNPs-NISM@BSA; 179.2 ± 6.7 nm for BiNPs-NISM-Pi3Ki@BSA; 188.9 ± 1.3 nm for BiNPs-NISM-Dox@BSA; 219.8 ± 6.3 nm for BiNPs-NISM-Dox/Pi3Ki@BSA; and 221.9 ± 3.5 and 221.7 ± 1.6 nm for the respective BiNPs-NISM-Dox/Pi3Ki@BSA-GSH and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems (Fig. 5A–C). Indeed, these findings further evidence that sequential modification of the nanosystem would bring about an increase in predictable hydrodynamic diameter, reflecting successive therapeutic agent (Dox and Pi3Ki) and targeting ligand (GSH and Glu) addition. From the bare 161.5 nm nanosystem, the total size went up to 221.7 nm for the encapsulation of dual drugs along with dual targeting. This shows that one could successfully assemble by layers, with each step contributing to the hydrodynamic radius. Notably, the size of the final formulation remains within the optimal range of 100–250 nm, which has been established as the optimum for penetration through the BBB. This size range represents a critical balance: particles below 10 nm are susceptible to rapid renal clearance, while those exceeding 250 nm face significant challenges in crossing the BBB and show preferential accumulation in the reticuloendothelial system. Evidence in the literature indicates that nanoparticles in the 50–200 nm range have improved permeability through the BBB via passive diffusion in glioblastoma regions of compromised BBB and receptor-mediated transcytosis⁵⁹. The zeta potential measurements confirmed unequivocally the necessary characteristics of colloidal stability of the formulated nanosystems. The zeta potential measurement value for BiNPs-NISM@BSA, BiNPs-NISM-Pi3Ki@BSA, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu were, respectively, −21.4 ± 0.6, −22.8 ± 0.9, −22.3 ± 0.7, −29.4 ± 0.4, −35.1 ± 1.0, and − 32.6 ± 0.4 mV (Fig. 5D–F).

Fig. 5.

Size and surface charge analysis of synthesized nanosystems by DLS. (A–C) Hydrodynamic diameter (average size in nm) of the synthesized nanosystems under optimal conditions, showing size distribution and uniformity. (D–F) Zeta potential (average in mV) of the synthesized nanosystems, indicating surface charge and colloidal stability.

These values reside within the range where nanoparticle formulations are assumed to have sufficient colloidal stability. According to established criteria, particles with zeta potential magnitudes exceeding ± 30 mV are considered electrostatically stable due to sufficient electrostatic repulsion forces that prevent aggregation⁶⁰.

Further, high and gradual surface charge accumulation resulting from GSH and Glu conjugations—from − 29.4 ± 0.4 mV (BiNPs-NISM-Dox/Pi3Ki@BSA) to −35.1 ± 1.0 mV (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH)—supports ligand inclusion and the thiol (-SH) and carboxyl (-COO^-) functionalization of the surface of the nanosystem with these functional groups on the surface of the nanosystem⁶¹. Moreover, the increased negative charge from using two ligands for functionalization was also from ionizable groups that are persistently deprotonated under physiological conditions: the cysteine thiol and carboxyl groups for GSH, as well as the α-carboxyl and γ-carboxyl groups for Glu^62,63. With this density of the surface charge, electrostatic stabilization is increased and makes it more effective in preventing aggregation by increasing repulsive forces between adjacent nanoparticles. The final formulation, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, has a zeta potential of −32.6 ± 0.4 mV, confirming its optimal characteristics for stability suitable for biological applications and for longer storage. The indices of polydispersity (PDI) for the synthesized nanosystems were determined to be as follows: for BiNPs-NISM@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, the numbers are 0.24 ± 0.04, 0.28 ± 0.13, 0.42 ± 0.06, 0.30 ± 0.08, 0.25 ± 0.02, and 0.27 ± 0.02, respectively. The PDI is a critical quality parameter that reflects the degree of homogeneity of a nanoparticle formulation with respect to size distribution.

According to international standards and pharmaceutical guidelines, a PDI value of between 0.01 and 0.4 indicates an acceptable size distribution; lower values below 0.3 are more desirable in lipid-based and polymer drug delivery systems. A PDI below 0.3 indicates a narrow, monodisperse size distribution^41,64.

Except for BiNPs-NISM-Dox@BSA (PDI = 0.42 ± 0.06), all other formulations showed PDI values below 0.3, which indicated excellent size distribution homogeneity. The only formulation with a slightly increased PDI (for Dox-encapsulated formulation) is indicative of the hydrophobicity of Dox and the tendency to form localized gradients during encapsulation, causing minor size heterogeneity.

More importantly, the final dual-drug, dual-targeted formulation (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu) displayed a PDI of 0.27 ± 0.02, thereby indicating that despite the complexity of the multi-component system, the process of formulation yielded a highly homogeneous nanoparticle population. This will be clinically very relevant since it indeed ensures consistent behavior for the whole nanoparticle population in terms of their BBB penetration efficiency, cellular uptake kinetics, drug release profiles, and therapeutic efficacy. In the present project, long-term stability studies have been tested for approximately two months at 4 °C. . It is worth noting that, validating the robustness of colloidal stability of the nanosystems under much longer storage conditions is crucial for future biomedical applications.

The complete physicochemical characterization indicates that the prepared nanosystems have maximum advantages for targeting glioblastoma. They are sized 161–222 nm, which puts them in the therapeutically relevant size range for BBB penetration since many preclinical studies indicate that nanoparticles in this size range have significantly better retention in the brain after administration through systemic use. The negative zeta potential ensures stability, minimizes nonspecific protein adsorption and complement activation, and reduces opsonization while improving the half-life in circulation⁶⁵.

Moreover, the narrow size distribution (PDI < 0.3) implies that almost all the nanoparticles depict similar biodistribution patterns, with similar efficacies in BBB translocation and kinetics of cellular internalization, both of which are critical for ensuring reproducibility in therapeutic outcomes and minimizing batch-to-batch variability. The successful incorporation of two therapeutic agents along with two targeting ligands while keeping these exceptional physicochemical parameters intact represents an advance toward accuracy in glioblastoma treatment through nanomedicine⁶⁶.

Long-term colloidal stability assessment of BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu

Considered a critical quality attribute, long-term physicochemical stability of nanocarrier formulations directly correlates to clinical translatability and shelf-life viability, ultimately translating into therapeutic efficacy. For the pharmaceutical nanoformulations designed for biomedical purposes, maintaining structural integrity and functional characteristics throughout extended storage time is important for reproducible therapeutic performance and regulatory acceptance. In this study, the long colloidal stability of the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu hybrid nanosystem was systematically assessed for eight weeks in a refrigerated condition (4 °C) by monitoring key physicochemical parameters, such as the hydrodynamic diameter, zeta potential, and polydispersity index. The hydrodynamic diameter measured as a function of time serves as a valuable indicator of physical stability and aggregation behavior of colloidal nanosystems. Observations of stability assessment reflected that the nanosystem BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu was treated with a size-stable formulation during an eight-week monitoring period (Fig. S1A).

Initially, the hydrodynamic diameter measured on day 1 at 221.7 ± 1.60 nm extended progressively to 237.9 ± 3.80 nm on day 28, 275.4 ± 15.10 nm on day 42, and 306.2 ± 4.10 nm on day 56, respectively, corresponding to cumulative increases in size variations of 7.3, 24.2, and 38.1%, against the initial measurement. Internationally accepted approaches for proving stability for lipid-based nanocarrier systems suggest that physically stable formulations must exhibit less than 15% size change, with the PDI value remaining under 0.3 for the duration of their testing⁶⁷.

Beyond a 15% size increase starting from day 28, thereby indicating gradual particle aggregation or vesicle fusion, was obviously beyond acceptable standards. Such an observed phenomenon might arise due to an interplay of several mechanisms operating in continued storage. First, the high degree of complexity of this nanosystem containing bismuth nanoparticles, two therapeutic agents, a BSA stabilizer, and two targeting ligands brings in multiple interfaces that are constantly rearranged due to inter-particle interactions⁶⁸. Notwithstanding the size increase, the very interesting point is that the formulation had maintained nanoscale characteristics throughout the storage times; this clearly suggests that final sizes are still in the therapeutically relevant range for glioblastoma applications since sizes were able to prevent aggregation. From statistical analysis, a significantly statistical time difference (***p < 0.001) appeared between day 1 and day 42 as a challenging high significance (****p < 0.0001) between day 1 and day 56. In contrast, the comparison between the consecutive times showed no significance (ns), such as between day 14 and day 28, day 28 and day 42, and day 42 and day 56. This result suggests an aggregation phenomenon occurring gradually over time rather than a sudden catastrophic destabilization. Interestingly, stability monitoring showed that during these eight weeks, PDI remained remarkably stable (Fig. S1B).

On day 1, the PDI was 0.27 ± 0.02, and since then, it had changed little to 0.33 ± 0.03 (day 14), 0.37 ± 0.09 (day 28), 0.45 ± 0.13 (day 42), and 0.42 ± 0.05 (day 56). The slight upward trend in PDI was noted after day 14 beyond the threshold of 0.3; however, the statistical analyses indicated no significant difference between all-time points (ns). This means that size distribution homogeneity was still reasonably maintained, an observation that is significant in light of the multi-component nature of the final nanosystem. The relative consistency of PDI value, despite size increase, indicates that the aggregation process is uniform for the entire nanoparticle population rather than affecting a subset of particles only. The slight PDI elevation witnessed at the later time points (day 42 and day 56) is indicative of the inherent challenges associated with long-term storage of complex multimodal nanocarrier systems. Lipid-based nanosystems are inherently dynamic structures that consistently reorganize under thermodynamic stability considerations. Along with such an observation, additional steric and electrostatic factors affecting colloidal stability arise due to hydrophilic targeting ligands GSH and Glu on the surface of the nanosystems. Lipid nanoparticle stability studies have reported similar PDI evolution patterns, where gradual increases in PDI are observed during prolonged storage and attributed to slow processes of particle ripening, lipid phase transitions, and surfactant redistribution⁶⁹.

Stability assessment also showed that zeta potential remained stable during the entire eight-week monitoring period (Fig. S1C), ranging from − 32.6 ± 0.4 mV on day 1 to −32.5 ± 0.9 mV (day 14), −28.2 ± 1.8 mV (day 28), −28.8 ± 0.742, and − 30.3 ± 2.4 mV (day 56). The slight reduction in the zeta potential magnitude value at day 28 (from − 32.5 to −28.2 mV) might reflect some surface reorganization process or slow charge screening effects. However, the following stabilization and partial recovery at the later time points (day 56: −30.3 mV) indicate that this electrostatic stabilization mechanism remained mainly intact.

The thorough stability investigations highlight that this BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem has an acceptable physicochemical stability within the first four weeks of storage at 4 °C in the fridge, with slow destabilization thereafter. Such a temporal stability profile can be correlated with behavior patterns commonly reported for lipid-nanocarrier systems, where multi-component formulations are more prone to long-term instability than their simpler counterparts. An 8‑week refrigerated (4 °C) stability assessment of particle size, PDI, and zeta potential is commonly reported for early nanoparticle formulation proof‑of‑concept studies and short‑term colloidal robustness evaluation in the literature.

In Altogether, this study performed an 8-week stability assessment at 4 °C, which is normal practice in early nanoparticle formulation proof-of-concept studies and short-term assessments of colloidal stability in the literature⁷¹. We recognize that translational and clinical development required extended real-time (≥ 6 months) and enhanced stability testing (size, PDI, zeta potential, and drug content). The absence of investigation into freeze-thaw stability and serum stability under physiological flow has been identified as a significant drawback of the stability study, requiring evaluation in future clinical studies.

Morphological and elemental characterization of nanocarriers

Transmission electron microscopy (TEM) and field emission scanning electron microscopy (FESEM) offered visualization of very high-scale resolution in the morphology and size distribution of the synthesized nanocarrier immediately before and after the incorporation of drugs and ligands. As represented in the TEM images (Fig. 6A,B), BiNPs-NISM@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanocarriers possessed a relatively smooth spherical architecture with uniform dispersion and limited aggregation: features of well-controlled synthesis and protein-mediated stabilization.

Fig. 6.

Morphological characterization and elemental analysis of synthesized nanosystems. (A, B) TEM images of BiNPs-NISM@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, showing particle size, shape, and dispersion. (C, D) FESEM images of BiNPs-NISM@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, illustrating surface morphology. (E, F) EDX analysis of BiNPs-NISM@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, confirming elemental composition of the nanosystems.

Most significantly, the effect of BSA as a stabilizer was instrumental in further supporting this regular spherical morphology, documented in the literature to ensure colloid stability and enhance biological interactions for nanocarrier platforms⁷². Additional analysis with FESEM (Fig. 6C,D) confirmed this morphological uniformity and maintained size measurements of nanoparticles both pre and post functionalized. Importantly, it is interesting that dual-drug, dual-ligand nanocarriers retained their spherical shape and size uniformity even after complex modifications, indicative of the technique being able to maintain nanosystem integrity throughout the process of multilayer assembly.

Energy-dispersive X-ray spectroscopy (EDS) (Fig. 6E,F), according to which the final nanocarrier formulation would be clearly evidence of elemental composition, showed such spectra revealing expected signature elements: carbon (C), nitrogen (N), oxygen (O), chloride (Cl), and significantly bismuth (Bi). Quantitative EDS data revealed the bismuth content to be approximately 5% by weight in the final nanoparticles, with carbon (50.89%), nitrogen (19.19%), and oxygen (24.84%) as constituents of the organic and NISM@BSA matrix, while chloride at very low concentration contributed insignificantly.

These qualities of strong presence of bismuth leave no doubt towards confirming that successful synthesis is achieved, incorporating metal nanoparticles into NISM nanocarriers - these, indeed, are requirements for nanoplatforms to qualify for radiosensitization and multimodal therapeutic delivery⁷³. The observed morphological and compositional characteristics are in line with what was gained from other advanced nanosystems studies, where optimal characteristics of biological behavior were characterized by spherical geometry, narrow size distributions, and robust inorganic–organic hybridization. As has been repeatedly demonstrated, this spherical morphology is also the condition that maximizes circulation half-life, cellular uptake, and tumor penetration while minimizing rapid clearance and off-target effects⁷⁴. The integration of both imaging modalities, TEM and FESEM, together with EDS evidence, establishes that not only does the final BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem contain a high payload incorporation, but it also possesses the capacity to maintain nanostructural and chemical integrity, the very basis of its encouraging biomedical promise.

Encapsulation efficiency, ligand conjugation, in vitro release, and kinetic modeling

The quantitative encapsulation efficiency of Dox and the Pi3Ki within the produced BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems was rigorously determined via UV-vis spectrophotometric calibration (Fig. S2). Specific linear relationships between absorbance (y) and concentration (x) were established for Dox at 495 nm, while Pi3Ki was at 294 nm for their respective spectra, thus enabling robust quantification. Calculations showed solid encapsulation efficiencies: 72.6 ± 5.0% for Dox and 68.5 ± 4.8% for Pi3Ki, consistent with or above previous reports for vesicular nanocarriers of similar complexity.

High encapsulation is indicative of the strong interactions and capacity of the NISM@BSA nanocarriers to stably entrap both hydrophilic and mildly hydrophobic compounds, which will support the optimum potential of this system for dual-drug delivery⁷⁵. Besides drug encapsulation, quantitative measurement of targeting ligand incorporation by indirect UV-vis analysis gave evidence for efficient covalent attachment of both GSH and Glu to the nanosystem. The conjugation efficiencies calculated 80.1 ± 6.2% for GSH and 78.3 ± 7.3% for Glu after repetition of measurement, the replication of successful surface functionalization.

Such high chemical binding yields are foundational for ensuring dual-targeted BBB transport and tumor specificity, thereby validating the synthetic approach and functional potential of the constructed nanosystems⁷⁶. The BiNPs-NISM@BSA nanocarriers exhibited significant drug entrapment levels, with Dox at approximately 72.6% and the Pi3K inhibitor approximately 68.5%. These values are comparable to or above several documented GBM nanocarriers. For example, targeted Dox encapsulation efficiency (EE) in liposomes has been reported at 90%⁸⁰, with typical liposomal systems showing EE of ≥ 90%⁷⁷, whereas PAMAM dendrimers exhibit a more moderate EE of approximately 59% Doxorubicin⁷⁸. Conversely, metallic nanoparticle systems often exhibit much reduced drug loading; Feng et al. reported approximately 16% doxorubicin encapsulation efficiency on fibroblast growth factor receptor-targeted gold nanoparticles. Consequently, our dual-drug payload nanoparticles integrate polymeric encapsulation efficiency with inorganic stability. Polymeric carriers, such as PEG-PTMC or PLGA nanoparticles, typically attain 90–95% encapsulation efficiency; however, our approximately 70% encapsulation efficiency is competitive, especially because of the complexity of co-encapsulating two agents⁷⁹.

Drug release profiling has shown highly pH-responsive release kinetics for the engineered nanosystem (Fig. 7A,B). At acidic pH (5.8) mimicking the tumor microenvironment and late endosomal/lysosomal compartments, cumulative release after 24 h reached 83.1% for Pi3Ki and 73.5% for Dox, while values were substantially lower (55.6% and 56.2%, respectively) at the physiological pH of 7.4. This presents a considerable improvement in the release at low pH, thus providing a key factor in tumor-targeted delivery since the destabilization of the nanocarrier membrane and cleavage of pH-labile linkages occur under acidic conditions.

Fig. 7.

Drug release and hemocompatibility analysis of synthesized nanosystems. (A) Release profile of Dox from BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems at pH 7.4 and 5.8, showing pH-dependent drug release behavior. (B) Release profile of Pi3Ki from BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems at pH 7.4 and 5.8, demonstrating controlled co-release of Pi3Ki. (C, D) Hemocompatibility analysis of BiNPs-NISM@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems at different concentrations, indicating their biocompatibility with red blood cells.

The pH responsiveness hence is comparable with earlier findings in innovative nanomedicine strategies and considerably uplifts prospects of site-selective chemotherapy in glioblastoma^80,81. Experimental release data were fitted to a suite of mathematical kinetic models (Table S2; Table 1) to understand the release mechanisms. At pH 5.8, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu release kinetics best fit the Gompertz model, with the highest correlation coefficients (R² = 0.9925) from the analyzed data, as well as minimum AIC and MSE values. At pH 7.4, the same model was best for free Dox, whereas alternative models such as Baker-Lonsdale and Weibull approached optimality for evaluated configurations of the system. The fact that the best fitting can be done with the Gompertz and related sigmoid or complex kinetic models suggests that the underlying release is governed by non-Fickian, multi-stage, or cooperative processes-that is, a combination that most likely and possibly reflects matrix diffusion, erosion, and pH-sensitive desorption.

Table 1.

R2, Akaike’s information criterion (AIC), and mean squared error (MSE) parameters from fitting release data of formulations on various models.

Formulation	BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu
pH	5.8			7.4
Model	R2	AIC	MSE	R2	AIC	MSE
Zero order	0.7714	125.70	1130.52	0.7630	120.75	772.52
First order	0.9726	102.27	186.44	0.9153	113.03	426.64
Higuchi	0.8960	110.27	344.96	0.8895	106.01	248.57
Korsmeyer-Peppas	0.9637	88.27	59.40	0.9686	79.66	30.63
Hixson-Crowell	0.9066	114.05	461.46	0.8810	115.92	532.97
Hopfenberg	0.9726	104.27	203.4	0.9154	115.03	465.49
Baker-Lonsdale	0.9535	96.48	119.48	0.9252	99.42	149.71
Makoid-Banakar	0.9840	79.75	29.10	0.9856	71.59	15.53
Peppas-Sahlin	0.9854	78.67	26.77	0.9876	69.76	13.49
Weibull	0.9907	72.51	16.67	0.9927	62.66	7.81
Quadratic	0.8710	118.45	605.31	0.8568	114.22	437.17
Logistic	0.9897	72.03	17.04	0.9857	69.48	14.00
Gompertz	0.9925	67.76	12.27	0.9929	60.44	6.99
Probit	0.9896	72.12	17.15	0.9869	68.30	12.78

The numbers shown in bold are actually the optimal numbers related to our prepared NPs formulation.

This modeling framework serves both for mechanistic understanding concerning controlled release design and aligns with the increasing consensus that hybrid, multi-ligand nanocarriers often follow nonclassical, multiphasic release patterns. Such release profiles are considered beneficial to coordinate drug availability with tumoral microenvironmental triggers, ultimately optimizing therapeutic efficacy⁸².

The synthesized nanosystems have suitable blood biocompatibility

Our results demonstrate favorable hemocompatibility for the drug-free BiNPs-NISM@BSA nanosystem. Hemolysis was negligible (≤ 0.80%) at concentrations up to 30 µg/mL and remained below the safety threshold of 5% (1.40–3.62%) even at the highest concentrations tested (45–90 µg/mL) (Fig. 7C). The lowest hemolysis (0.84% at 5 µg/mL) was observed for this formulation, indicating minimal red blood cell (RBC) damage. In contrast, the drug-encapsulated BiNPs-NISM-Dox/Pi3Ki@BSA-GSH/Glu group showed an apparent concentration-dependent increase in hemolysis (Fig. 7D). While hemolysis was low at 5 µg/mL (0.84%) and 30 µg/mL (2.35%), it exceeded the common 5% safety threshold at higher doses (5.56% at 45 µg/mL and 22.58% at 90 µg/mL). This difference confirms the key role of the encapsulated drugs (Dox/Pi3Ki) in increasing toxicity, likely due to their inherent cytotoxic effects, alterations in nanoparticle surface properties, or direct interaction between the drugs and the RBC membrane. Regulatory and scientific standards for acceptable hemolysis vary (5% per ASTM E2524-08⁸³, ≤ 2% per American Society of Clinical Pathology^84,85. Based on the stringent 5% threshold, the drug-loaded nanocarriers are safe only at concentrations ≤ 30 µg/mL. Therefore, while the drug-free system is hemocompatible at all tested concentrations, the drug-loaded system requires dose optimization (≤ 30 µg/mL) or surface modifications to reduce hemolytic risks.

Cell uptake of the synthesized nanosystems

The efficiency of cellular internalization is a key determinant in evaluating the potential of nanocarriers for targeted drug delivery. In the present study, the cellular uptake pattern of the synthesized nanosystems was determined quantitatively in U87 glioblastoma and HEK-293 cells by using flow cytometry (Fig. 8). Obtained results from Ctrl group exhibited minimal fluorescent signals (2.13 ± 0.82%) which revealed organelles related fluorescence inside the cell, while BiNPs-NISM@BSA (10 µg/mL) showed a slight increase (3.16 ± 1.06%) indicating that the components of the synthesized nanosystem are non-fluorescent. On the other hand, BiNPs-NISM-Dox/Pi3Ki@BSA presented significantly higher cellular internalization (73.90 ± 5.51%) at the same concentration. The dual-targeted formulation, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu, exhibited an apparent dose-dependent increase in internalization, achieving uptake efficiencies of 26.20 ± 3.43%, 74.20 ± 4.19%, 93.70 ± 2.57%, and 94.60 ± 3.02% at the concentrations of 1, 2.5, 5, and 10 µg/mL, respectively. The cell uptake rate in the HEK-293 cell group was much lower at 10 µg/mL (62.30 ± 3.30%), possibly due to the reduced expression of glutamate receptors in HEK-293 cells, thereby establishing the specificity of the glutamate-mediated targeting strategy⁸⁶. Statistically, it was revealed that all treatment groups (except BiNPs-NISM@BSA treated group) compared to Ctrl were different significantly (p < 0.0001), in which the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu (10 µg/mL) exhibited the highest uptake (Fig. 8A,B).

Fig. 8.

Cellular uptake of synthesized nanosystems in U87 and HEK-293 cells. Cellular uptake of BiNPs-NISM@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems in U87 and HEK-293 cell lines, measured at different concentrations. Statistical analysis of uptake rates is shown, with significance indicated as ****p < 0.0001 (n = 3).

HEK-293 cells served as non-tumor controls due to their established absence of endogenous glutamate receptor expression. Collett et al. showed that HEK-293 cells exhibited low endogenous glutamate receptor activity using calcium imaging with fluo3-AM in response to NMDA and DHPG, revealing practically no response⁸⁶, thus facilitating a dependable evaluation of targeting specificity relative to U87 GBM cells.

The observed uptake behavior represents a crucial finding, as it identifies an optimal concentration at which the nanosystems achieve maximal internalization without necessitating higher doses that could raise cytotoxicity concerns. The enhanced cellular uptake of the nanosystem can be associated with its small size to facilitate active and passive internalization⁸⁷ as well as to ligand-mediated targeting that enables receptor-specific uptake³⁹. Previous studies have shown that the modification of nanoparticle surfaces significantly affects internalization efficiency. Gao et al. revealed that nanoparticles coupled with the IL-13 ligand exhibited enhanced internalization by glioma cells⁸⁸. Previous research demonstrated that glutamate-modified liposomes significantly enhanced the cellular uptake and cytotoxic efficacy of docetaxel in C6 glioma cells compared to unmodified liposomes, indicating the important function of ligand conjugation in targeted cellular delivery^89,90. The current findings confirm that the surface functionalization of BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu with glutathione and glutamate facilitates receptor-mediated uptake in glioblastoma cells while maintaining selectivity for non-cancerous cells, such as HEK-293 cells.

The present study does not experimentally assess transport across the BBB, although showing that the GSH/Glu–functionalized BiNPs-NISM nanosystem achieves effective and selective uptake in U87 glioblastoma cells. The dual-targeted design was rationally based on GSH recognition by Na-dependent glutathione transporters enriched on brain endothelium and Glu/AMPA-receptor-mediated uptake in glioblastoma cells. Nevertheless, the available results are limited to direct tumor-cell experiments and should be interpreted as an in vitro tumor-level proof-of-concept rather than a comprehensive demonstration of brain delivery. In order to directly evaluate the trans-endothelial migration of GSH-functionalized versus non-targeted nanoparticles, we are aiming to develop validated in vitro BBB models (such as endothelial–astrocyte/pericyte co-culture systems on Transwell inserts with TEER and tight-junction verification). To completely validate the dual-targeting strategy for brain administration, these BBB-focused investigations will be followed by in vivo biodistribution and brain/tumor accumulation analysis in orthotopic glioma models.

The synthesized nanosystems reduce cell viability

In the No X-irradiation condition (Fig. 9A), all groups except for BiNPs-NISM@BSA showed a significant decrease in cell viability compared to the control group (p < 0.0001). Furthermore, no significant differences were observed among any of the concentrations of BiNPs-NISM@BSA group.

Fig. 9.

Cell viability and gene expression analysis after treatment with nanosystems. (A, B) Cell viability of U87 cells treated with different nanosystem groups including BiNPs-NISM@BSA, Free Dox, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu at concentrations of 2.5, 5, and 10 µg/mL, measured (A) without X-ray exposure and (B) with X-ray exposure. (C) qRT-PCR evaluation of Free Dox, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu groups at 5 µg/mL, showing gene expression changes. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n = 3).

In the Free Dox group, all concentrations differed significantly from one another (p < 0.0001). Similarly, for the BiNPs-NISM-Dox@BSA group, cell viability at concentrations of 5 and 10 µg/mL was significantly reduced compared to 2.5 µg/mL. In the cell group treated with the BiNPs-NISM-Pi3Ki@BSA, significant differences in viability were observed at 2.5 and 5 µg/mL when compared to 10 µg/mL. Regarding the BiNPs-NISM-Dox/Pi3Ki@BSA group, a significant decrease in cell viability was noted at 2.5 µg/mL compared to both 5 µg/mL (p < 0.01) and 10 µg/mL (p < 0.0001). In the cell group treated with the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, a significant decrease was observed from 2.5 to 5 µg/mL (p < 0.1), and from 5 to 10 µg/mL (p < 0.0001). Treatment with BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu decreased cell viability significantly at 5 and 10 µg/mL compared to 2.5 µg/mL (p < 0.001 and p < 0.0001, respectively). In this group, a significant reduction was also recorded between 5 and 10 µg/mL (p < 0.01). Notably, this group exhibited significantly lower cell viability at 5 µg/mL compared to the BiNPs-NISM-Dox/Pi3Ki@BSA group at the same concentration (p < 0.01). A similar trend was observed at 2.5 µg/mL (p < 0.1). Moreover, at 5 µg/mL, this group showed significantly reduced cell viability relative to the Free Dox group at the same concentration (p < 0.0001).

Under X-irradiation conditions (Fig. 9B), all groups (except BiNPs-NISM@BSA at 2.5 µg/mL) showed a significant decrease in cell viability compared to the control group (p < 0.0001). In the Free Dox group, all concentrations differed significantly in cell viability reduction from each other (p < 0.0001), and overall cell viability decreased compared to non-irradiated conditions. Cell-treated with the BiNPs-NISM@BSA-Dox exhibited a response pattern similar to that under non-irradiated conditions, though with a further reduced survival rate. Cell viability rate in the BiNPs-NISM@BSA-Pi3Ki group at all concentrations differed significantly from one another (p < 0.0001), and survival rates were lower than under non-irradiated conditions.

In the BiNPs-NISM-Dox/Pi3Ki@BSA group, specifically, cell viability at 2.5 µg/mL was significantly lower than at 5 µg/mL (p < 0.01) and 10 µg/mL (p < 0.0001), and a decrease was also observed at 10 µg/mL compared to 5 µg/mL (p < 0.1). For the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH group, a significant decrease in cell viability occurred at 10 µg/mL in comparison to both 2.5 and 5 µg/mL (p < 0.001). Altogether, in the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu group, cell viability decreased significantly at 5 and 10 µg/mL compared to 2.5 µg/mL (p < 0.1 and p < 0.0001, respectively), and a significant reduction was also observed at 10 µg/mL relative to 5 µg/mL (p < 0.0001). The obtained results after treatment with 2.5, 5, and 10 µg/mL of various nanosystems in No X-irradiation condition showed that the IC50 value for the Free Dox drug, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu group were 8.12 ± 0.83, 6.65 ± 1.46, 3.45 ± 1.02, 8.08 ± 0.79, and 3.33 ± 0.20 µg/mL, respectively (Fig. S3A). In 2 Gy X-ray exposure condition, the IC50 value for the Free Dox drug, BiNPs-NISM-Dox@BSA, BiNPs-NISM-Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu treated group were 6.60 ± 1.05, 4.62 ± 0.31, 5.94 ± 0.61, 2.12 ± 0.11, 2.67 ± 0.36, and 1.58 ± 0.06 µg/mL, respectively (Fig. S3B). Based on the obtained IC50 value for the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu in the both condition, concentration of 5 µg/mL to further assays was choose. As mentioned in the results, our dual-targeted nanosystem (BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu) exhibited significantly higher U87 cell cytotoxicity than free Dox or singly-targeted particles which is similar to the previous reports. Regarding the com bination effect, El Moutaoukil et al. (2025)⁹¹ demonstrated that dual-functional AuNPs (Dox + neurofilament peptide) significantly enhanced GBM cell death compared to mono-functionalized controls.

In addition, Marklein et al. reported that inhibition of the PI3K pathway potentiates Dox-triggered apoptosis in sarcoma cells¹⁶, supporting the enhanced GBM chemosensitivity achieved by the PI3K inhibitor used in the present study.

Despite the dose-dependent cytotoxicity of Free Dox drug, it does not provided vagarious effect compared to other formulated particles that indicates the low efficacy of Dox on U87 cells, which is due to intrinsic chemoresistance mechanisms⁹² and the activity of the efflux pumps⁹³. On the other hand, encapsulation of Dox in hybrid nanocarrier (BiNPs-NISM-Dox@BSA) greatly enhanced cytotoxicity that can lead to improve intracellular accumulation and prolonging retention. This effect was enhanced under irradiation conditions, which supports previous studies showing that Dox can stabilize DNA damage by generating reactive oxygen species (ROS) and acts as a radiosensitizer⁶. Treatments with BiNPs-NISM-Pi3Ki@BSA indicate LY294002’s ability to inhibit Pi3K/Akt signaling in glioblastoma⁹⁴. This confirmed the role of Pi3K/AKt inhibition in suppressing glioblastoma survival pathways. Activation of this pathway is one of the most common oncogenic changes in GBM. In a study conducted by Li et al. it was shown that inhibition of the Pi3K pathway using LY294002 can sensitize U87 cancer cells to radiation by preventing DNA damage repair and inhibiting the AKt signaling pathway, which promotes cell survival⁹⁵. Co-encapsulation of Dox and Pi3Ki in BiNPs-NISM-Dox/Pi3Ki@BSA group exhibited greater cytotoxicity in agreement with a synergistic effect in keeping with Pi3K inhibition sensitizing tumor cells to DNA-damaging agents¹⁶. It could be due to the presence of LY294002 and its ability to increase the sensitivity of cancer cells to chemotherapy drugs such as doxorubicin¹⁶.

The reduced U87 cell viability observed under X-ray irradiation, attributed to Bi-mediated radiosensitization, is further supported by recent studies, as He et al. reported increased GBM cell death following treatment with angiopep-2/Au-DOX combined with radiotherapy⁹⁶. Moreover, Kang et al. showed that Bi-based Cu₃BiS₃@ZIF8@Dox system achieving ~ 87.6% tumor suppression under X-ray⁹⁷. Collectively, these studies confirm the strong potential of Bi-based nanomaterials in inducing GBM cell death, consistent with our findings. Notably, Bi nanoparticles exhibit superior radiosensitization compared with other metal nanoparticles, as Hossain et al. found that nano-Bi provided ~ 1.25× the radiosensitization than Au and ~ 1.29× that of Pt at similar doses⁹⁸.

Functionalization of the synthesized nanosystem surface with GSH reduced cell viability compared to controls. However, glutathione plays a paradoxical role in glioblastoma biology, as high intracellular glutathione levels can support tumor growth and therapy resistance^20,99. Although GSH facilitates nanoparticle transport across the blood–brain barrier¹⁹, its antioxidant and detoxifying activities can promote treatment resistance in cancer cells^99,100; therefore, we functionalized our nanosystem with GSH to better mimic physiological delivery to assess its impact on in vitro cytotoxicity and to determine whether GSH functionalization should be pursued in future in vivo studies. Nevertheless, dual targeting with GSH and Glu produced the most potent cytotoxic effect in all conditions, indicating that the use of Glu ligand enhanced uptake mediated by pre-expressed glutamate receptors on U87 cancer cells¹⁰¹. However, we did not find any studies regarding the interference of Glu with the antioxidant and detoxifying of GSH, that suggested to be evaluated in future. Despite the biocompatibility of BiNPs-NISM@BSA in No X-irradiation treatment, under X-ray exposure, cell viability decreased compared to conditions without X-irradiation.

This could be due to the presence of bismuth and its radiosensitizer properties, which due to its high atomic number, can absorb a lot of radiation at that point, leading to elevated DNA damage under irradiation and thus radiosensitization of tumor cells¹⁰². Free Dox drug was also reducing viability compared to the No X-irradiation treatment, which can confirm that doxorubicin is a radiosensitizer¹⁰³. In the BiNPs-NISM-Dox/Pi3Ki@BSA group, the survival rate was reduced compared to conditions without X-irradiation, highlighting the radiosensitizing and chemosensitizing roles of Pi3K pathway inhibition in GBM models, as established in studies where Pi3K blockade increases tumor sensitivity to DNA-damaging^104,105.

The synthesized nanosystems downregulated cancerous properties related genes

According to the qRT-PCR data (Fig. 9C), U87 cell treatment with BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu group significantly decreased the expression levels of Vimentin, Bcl-xL and Cyclin D1(fold change: 0.331 ± 0.13, p-value: <0.0001; fold change: 0.456 ± 0.074, p-value: <0.0001 and fold change: 0.323 ± 0.141, p-value: <0.0001; respectively) compared to the control group. These results are in accordance with cell viability results. As our MTT significantly represented decreased U87-cell viability in the control group compare to the BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu group, it could be associated to the downregulation of these genes. Vimentin is a key filamentous protein that plays a fundamental role in the structure and function of cells. Vimentin concentration is low during the early stages of cancer development, but it increases as the cancer invasion to the surrounding areas¹⁰⁶, thereby its downregulation could interfere with cancer cell migration, which are in accordance with cell scratch evaluation (Fig. 12). Moreover, the antiapoptotic Bcl-xL, have a central role in promote cancer cell survival and metastasis^107,108, as we observed in apoptosis evaluation (Fig. 11), the decreased expression can lead to apoptosis in cancer cell, which finally can affect cancer cell viability. Previous studies separately confirmed the Bcl-xL downregulation in cancer cells treated with Dox drug or LY294002^109,110, which are consistent with our results. Cyclin D1 is a crucial cell cycle regulator that plays a critical role in tumorigenesis through the induction of uncontrolled cell proliferation. While its expression is tightly controlled in normal cells, cancer cells exhibit higher Cyclin D1 activity¹¹¹. Its downregulation in our study could be attributed to the presence of Pi3Ki and Dox drug. However, Dox has shown contradictory results in previous studies; as Dox treatment MDA-MB-231/ADR cells reduce cyclin D1 expression¹¹², Gopinath et al. showed that Dox did not represented considerable effect on Cyclin D1 expression¹¹³.

Fig. 12.

Cell migration inhibition after treatment with different nanosystems. U87 cell migration was evaluated after treatment with 2.5 μg/mL of BiNPs-NISM@BSA, Free Dox, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems under the following conditions: (A) pre-treatment (0 h), (B) without X-ray exposure (48 h), and (C) with X-ray exposure (48 h). (D) Statistical analysis shows the percentage of cell migration inhibition. Statistical significance: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (n=3).

Fig. 11.

Cell apoptosis analysis after treatment with different nanosystems. Apoptosis evaluation of U87 cells treated with BiNPs-NISM@BSA, Free Dox, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu at 5 µg/mL under the following conditions: (A, C) without X-ray exposure and (B, D) with X-ray exposure. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n = 3).

On the other hand, LY294002-treated colon cancer stem cells showed decreases in expression of Cyclin D1, which confirms our results¹¹⁴. Moreover, Pi3Ki have reported to affect Cyclin D1 expression through various pathways including reactivates GSK-3β, which phosphorylates Cyclin D1 on Thr286 and promotes^115,116 and nuclear FOXO activity that represses D-type cyclin transcription^117,118. Moreover, Pi3K–Akt also upregulates Cyclin D1 through CREB-dependent transcriptional control¹¹⁹. Therefore, it can be concluded that Pi3Ki embedded in our nanosystem may have the most effective role to obtain such a result.

The synthesized nanosystems arrested the cell cycle

In the No X-irradiation condition (Fig. 10A,C), cells arrested in the sub-G1phase after treatment with 5 µg/mL of BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems and the control group (Ctrl) were 3.91 ± 0.66, 6.18 ± 2.21, 19.98 ± 3.27, 43.86 ± 3.12, and 2.39 ± 0.59%, respectively. Other specific cell cycle arrest phases are in shown in Fig. 10.

Fig. 10.

Cell cycle analysis after treatment with different nanosystems. Cell cycle evaluation of U87 cells treated with BiNPs-NISM@BSA, Free Dox, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu at 5 μg/mL under the following conditions: (A,C) without X-ray exposure and (B, D) with X-ray exposure. Statistical significance: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (n=3).

Under X-irradiation (Fig. 10B,D), treatment with BiNPs-NISM-Dox/Pi3Ki@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu showed predominant cell arrest in the sub-G1 phase as 25.71 ± 2.36 and 51.71 ± 3.59, respectively, that were remarkably distinct from those observed in Ctrl, BiNPs-NISM@BSA, and Free Dox drug (4.79 ± 1.20, 4.77 ± 1.45, and 8.94 ± 1.61, respectively). These results emphasize the potential of this nanosystem in enhancing the therapeutic effects of X-irradiation as a radiosensitizer. As expected, the BiNPs-NISM@BSA nanocarrier did not significantly alter the cell cycle pattern compared to the Ctrl, reflecting the biocompatibility of this nanocarrier.

However, treatment of cells with Free Dox drug along with both X-irradiation condition induced cells to arrest mostly in S-phase compared to the Ctrl group, consistent with the mechanism of Dox, which is a DNA intercalator and topoisomerase II inhibitor that can disrupt DNA replication⁶ and in line with our previous study that have reported the S-phase arrest of SW480 Cell line in Dox treatment group compared to Ctrl group¹²⁰. However, the low sub-G1 population cells in Dox treatment group are associated to the Dox limitation to overcome efflux pumps. Glioblastoma cells are relatively resistant to Dox due to overexpression of ATP-binding cassette and P-gp efflux transporter⁹³. BiNPs-NISM-Dox/Pi3Ki@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems significantly increased the sub-G1 population cells compared to other groups, while it was more substantially increased in the dual-targeted formulation. These interesting findings could be attributed to several mechanisms, including preventing Dox drug efflux through encapsulation within the NISM. Xu et al. also reported that loading of Dox in Poly(D, L-lactide-co-glycolide) (PLGA) decreased the efflux rates of Dox lading to more cell death compared to Free Dox drug¹²¹, confirmed the more effect of our formulation by encapsulating Dox drug. Pi3Ki, on the other hand can interfere with cell cycle process. Inhibition of Pi3K suppresses growth-promoting signaling and enhances Cdk inhibitor–induced apoptosis, thereby halting cell cycle progression and reactivating tumor-suppressive mechanisms¹²². Moreover, the results of a previous study showed that adenovirally infected C81 cells treated with LY294002 can increase the number of cells in the Sub-G1 phase¹²³. Taking together, our formulation increased the sub-G1 population through the synergistic effect of Dox prolong cell habitation, the Pi3Ki activity and cell specific targeting through GSH and Glu.

Another exciting result was the enhanced cell cycle arrest under X-irradiation through more sub-G1 arrest which is due to the presence of the radiosensitizers components (related to the radiosensitizing properties of bismuth) in this synthesized nanosystems, as discussed in previous Sect¹²⁴.

The synthesized nanosystems induced cell apoptosis

In the No X-irradiation condition (Fig. 11A,C), total apoptosis rate (early and late) after treatment with 5 µg/mL of BiNPs-NISM @BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems and the control group (Ctrl) were 8.06 ± 3.28, 17.15 ± 2.27, 31.52 ± 412, 49.25 ± 3.62, and 5.78 ± 1.43%, respectively. The obtained results showed that the apoptosis rate in all treated groups, except the BiNPs-NISM@BSA group, increased compared to the control (untreated) group. The apoptosis rate induced in the cell groups treated with BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem was significantly higher than in other treated groups.

Under X-irradiation condition (Fig. 11B,D), cell-group treated with 5 µg/mL of BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems led to induce total apoptosis rate equal to 18.27 ± 2.85, 29.42 ± 3.87, 41.02 ± 3.84 and 65.40 ± 3.24, respectively, while it was 9.05 ± 1.95% in control group. All treated groups with different nanosystems have shown a significant apoptosis rate compared to the control group. The cells treated with BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu compared to the BiNPs-NISM @BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA groups showed significantly more apoptosis (p < 0.0001).

The apoptosis results demonstrated that treatment with Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems produced a significant pro-apoptotic effect in U87 cells, which was enhanced by X-irradiation. In the No X-irradiation condition, BiNPs-NISM@BSA induced a slight increase in apoptosis compared to the control, confirming its biocompatibility.

Free Dox drug induced moderate apoptosis in treated cell group, consistent with previous report of its limited efficacy on U87 cells due to multidrug resistance mechanisms such as p-glycoprotein efflux⁹³. The moderate effect of Free Dox drug compared to nanoformulated Dox on apoptosis has been well documented in previous studies^41,120. However, in our research co-encapsulation of Dox drug with the Pi3Ki significantly increased apoptosis, demonstrated the critical role of Pi3K pathway inhibition in sensitizing glioblastoma cells to DNA damaging agents, which is consistent with Marklein et al. who reported that Pi3K inhibition enhances doxorubicin-induced apoptosis in sarcoma cells¹⁶. Moreover, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem induced the highest level of apoptosis, indicating that the use of dual ligand targeting can also increase apoptosis by increasing intracellular uptake. Glu and GSH receptors are highly expressed in glioblastoma cell lines^18,19. Collectively these studies, confirm the fact that targeting with GSH and Glu could specifically target U87 cell lines as well documented in our study, which finally led to high accumulation inside the cellular environment that may result in higher rate of apoptosis as we observed. To confirms our apoptosis, we must mention the Real-time PCR results that indicated that treated cells with BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem showed more substantially decrease in the Bcl-xL gene expression, which as an antiapoptotic gene, its reduction can affect the apoptosis elevation in U87 cells. Furthermore, we observed more increase apoptosis rate in all groups after X-irradiation treatment. This important is associated with the present of bismuth metal¹⁰² which act as a key radiosensitizer. However, it should mention that Dox¹⁰³, and Pi3Ki^104,105 could act as radiosensitizer agents that could further sensitized U87 cell line to X-irradiation.

In our study, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu treatment led to pronounced apoptosis, sub-G1 cell-cycle arrest, and downregulation of Bcl-xL and Cyclin D1 in U87 cells, consistent with engagement of apoptotic and anti-proliferative pathways. Nevertheless, because we did not directly measure ROS production or caspase activation, the precise molecular cascade (e.g., ROS-driven mitochondrial apoptosis versus predominantly Pi3K/Akt-dependent signaling blockade) remains to be clarified. Future work must therefore incorporate ROS assays, mitochondrial membrane potential analysis, and caspase-3/9 activity measurements to mechanistically align our radiosensitizing chemo-/Pi3K-inhibitory platform with the well-characterized apoptosis pathways reported for other advanced nanotherapeutics. Although Annexin V/PI flow cytometry and down-regulation of the anti-apoptotic marker Bcl-xL suggested apoptosis, a limitation of this work is that executioner caspase-3/7 activity was not directly measured. Therefore, to more accurately identify the apoptotic pathways activated by the dual-targeted nanosystem, future research will consist of caspase-3/7 assays and extended signaling analysis.

The synthesized nanosystems inhibited cell migration

In No X-irradiation condition (Fig. 12B,D), the control group showed minimal cell migration inhibition (3.17 ± 1.58%), reflecting the high intrinsic motility of U87 cells while 48 h after cell treatment with the BiNPs-NISM@BSA nanocarrier was represented non-significant cell migration inhibition than the control group (5.41 ± 1.53%).

On the other hand, the Free Dox drug showed a higher rate of cell migration inhibition (12 ± 2.25%) which were even more enhanced by the BiNPs-NISM-Dox/Pi3Ki@BSA treatment that reached to 18.66 ± 1.56%, indicated the simultaneous presence of the Pi3Ki and Dox drug. The most pronounced effect (26.95 ± 2.58%) was observed with dual-targeting BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystem. Under X-irradiation conditions (Fig. 12C,D), the cell migration inhibition for control, BiNPs-NISM@BSA, Free Dox drug, BiNPs-NISM-Dox/Pi3Ki@BSA, BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu were 4.91 ± 1.49%, 9.33 ± 1.94%, 15.47 ± 2.50%, 33.37 ± 3.02%, and 46.16 ± 3.41%, respectively. Altogether, X-irradiation induced more substantial cell migration inhibition in all treated groups.

The cell migration inhibition results further highlight the therapeutic potential of the developed nanosystems against the invasive phenotype of glioblastoma cells¹²⁵. Interfering with cytoskeletal dynamics¹²⁶ and inducing DNA damage¹²⁷ can be associated to the moderate migration effect of Dox that can led to reduce glioma motility. Interesting it should be noted that the secretion of matrix metalloproteinases (MMPs) following by the promotion of Pi3K/Akt led to the aggressive nature of U87 cells¹²⁸. Inhibiting Pi3K/AKt signaling pathway reduced actin cytoskeleton remodeling and MMP expression, thereby limiting invasive capacity^129,130. Here we noted that U87 cells treatment with BiNPs-NISM-Dox/Pi3Ki@BSA and BiNPs-NISM-Dox/Pi3Ki@BSA-GSH-Glu nanosystems showed higher rate migration inhibition, represented the synergic dual action of Dox-induced DNA damage and Pi3K inhibition and also remarked the importance of targeting for elevation the treatment efficacy.

Conclusion

In the present study, successful synthesis of an innovative nanosystems that be co-encapsulated with Dox and Pi3Ki and is dual-targeted by glutathione and glutamate ligands were confirmed using various characterization assays. The biological investigations conducted on the U87 glioblastoma cells showed that this multi-functional nanosystem exerted distinct anticancer effects through the significant cell viability reduction, gene expression downregulation, cell cycle arrest, apoptosis induction, and cell migration inhibition. All antitumor measures were, however, significantly more enhanced by the final formulation co-loaded with both therapeutic agents and dual GSH/Glu-targeting. Mechanistically, the bismuth constituent radiosensitized tumor cells to irradiation, whereas the Pi3K inhibitor elevated doxorubicin cytotoxicity through the inhibition of survival signaling. Moreover, glutamate targeting might inhibit any potential growth promotion from glutathione, increasing specificity and enhancing the overall therapeutic window, which requires further investigation in future studies. Furthermore, under X-irradiation, these synergistic modalities would amplify tumor cell response and biological consequences. Indeed, the combination of radiosensitizer, signaling pathway inhibitor, and chemotherapeutic agents in a dual-ligand targeted delivery nanocarrier would be highly applicable against the therapeutic resistance observed in glioblastoma treatment. These reported findings pave the way for further evaluation and possible translational development of this nanosystem as a sophisticated precision-engineered platform for combinatorial cancer therapy in this quite advanced disease.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BiNPs

Bismuth nanoparticles

NISM

Niosome nanoparticles

FT-IR

Fourier transform infrared

UV-vis

Ultraviolet–visible

PI

Propidium iodide

BSA

Bovin serum albumin

Dox

Doxorubicin drug

Pi3Ki

Pi3Kinase inhibitor

FE-SEM

Field emission-scanning electron microscopy

TEM

Transmission electron microscopy

GSH

Glutathione

Glu

Glutamate

EDS

Energy-dispersive X-ray spectroscopy

PDI

Polydispersity index

IC50

Half-maximal inhibitory concentration

pH

Potential of hydrogen

DMSO

Dimethyl sulfoxide

MSE

Mean squared error

AIC

Akaike’s information criterion

Author contributions

**Z.B.:** Methodology, Software, Writing—original draft. **M.G.:** Conceptualization, Methodology, review, and editing. **G.B.:** Methodology. **H.R.** : Methodology. **B.J.** : Supervision, Conceptualization, Methodology, Project administration, Writing—review, and editing.

Funding

This work is based upon research funded by the Iran National Science Foundation (INSF) under project No.4028565 and also supported by Zanjan University of Medical Sciences, Zanjan, Iran (Grant Number: A-12-1244-28 & Ethical Code: IR.ZUMS.BLC.1402.027).

Data availability

The data sets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable. U87 human glioblastoma cells (ATCC HTB-14) were purchased from the Pasteur Institute of Iran (Tehran, Iran). The hemocompatibility assay involving human red blood cells was approved by the Ethics Committee of Zanjan University of Medical Sciences, Zanjan, Iran (Ethical Code: IR.ZUMS.BLC.1402.027), and all procedures were performed in accordance with the institutional guidelines and the Declaration of Helsinki. Written informed consent was obtained from the donor before blood collection.

Consent for publication

Written informed consent for publication of anonymized data was obtained from the blood donor.