Polymer-stabilized amorphous CuO–ZnO hybrid nanocomplex as a promising candidate for antimicrobial therapy and controlled drug delivery with molecular docking insights

Abstract

In this study, multifunctional polymer-stabilized CuO–ZnO hybrid nanocomplexes were developed through a green synthesis approach and systematically evaluated for their antimicrobial and protein delivery potential. Bimetallic CuO–ZnO nanoparticles were biosynthesized using extracts of Cotoneaster horizontalis, Salvia officinalis, and Laurus nobilis, followed by integration into chitosan/nanochitosan–P(MMA-co-MAA) copolymer matrices. The resulting nanocomplexes were characterized using SEM, DLS, FT-IR, and XRD analyses, confirming successful nanoparticle formation, homogeneous dispersion, and predominantly amorphous structures, with particle sizes ranging from 62 to 138 nm. Among the formulations, nanochitosan-based systems exhibited improved structural compactness and reduced polydispersity. Molecular docking simulations revealed strong and stable interactions between the Cu–Zn alloy-modified polymeric carriers and bovine serum albumin, as well as remarkably high binding affinities toward key bacterial and fungal virulence proteins, including Sortase A, TolC, and CYP51. These findings suggested enhanced multivalent interaction capabilities of the nanocomplexes. In vitro antimicrobial assays corroborated the computational predictions, demonstrating pronounced antibacterial and antifungal activities (15–22 mm against E. coli, 17–24 mm against S. aureus and 21–25 mm against A. niger) with low minimum inhibitory concentrations (0.50 mg/mL for E. coli and 0.25 mg/mL for S. aureus). Furthermore, in vitro release studies using bovine serum albumin and human insulin as model biomolecules revealed sustained and controlled release profiles over seven days, with minimal burst effects. Nanochitosan-based systems exhibited slightly slower release kinetics, attributed to denser polymeric networks and stronger biomolecule–carrier interactions. The incorporation of P(MMA-co-MAA) further contributed to pH-responsive and diffusion-controlled release behavior. Collectively, this study presents an integrated experimental and computational strategy for designing biocompatible, green-synthesized CuO–ZnO nanocomplexes with dual antimicrobial and drug delivery functionalities. The developed platform may hold strong potential for biomedical applications, including drug delivery systems, wound dressings, and implant-associated infection control.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-57132-x.

Introduction

Polymer-stabilized nanocomplex-based controlled drug release systems are an innovative approach to enhancing the efficiency and precision of drug delivery. These systems integrate nanoparticles with polymeric or inorganic matrices to create multifunctional platforms capable of protecting drugs, targeting specific tissues, and releasing their payloads in a controlled manner. By tailoring the properties of the polymer-stabilized nanocomplexes, such as size, surface chemistry, and responsiveness to stimuli (pH, temperature, or light), researchers have improved drug stability, bioavailability, and therapeutic efficacy while reducing side effects. Applications range from cancer therapy and antimicrobial treatments to managing chronic diseases¹. Recent advancements emphasize size-controlled polymer-stabilized nanocomplexes, such as sub-100 nm carriers developed via template polymerization, which enable precise control over drug release and delivery. These systems are particularly valuable for delivering sensitive biomolecules like DNA or proteins, as they maintain stability while ensuring effective delivery to targeted cells². Polymer-stabilized nanocomplexes containing bimetallic nanoparticles are increasingly being developed for drug delivery systems due to their unique properties. These systems leverage the synergistic effects of two different metals, such as gold and silver or platinum and palladium, to enhance properties like stability, catalytic activity, and therapeutic efficacy. These bimetallic nanoparticles are often embedded in polymeric or inorganic matrices, improving drug loading, release control, and biocompatibility. Such innovations show potential in cancer treatment and antimicrobial therapies³. Bimetallic nanoparticles are composed of two distinct metals, offering enhanced properties due to synergistic effects between the metals. These nanoparticles exhibit improved stability, catalytic activity, biocompatibility, and reduced toxicity compared to their monometallic counterparts. Such properties make them valuable in biomedical applications, including drug delivery, imaging, and therapy. The key medical applications of bimetallic nanoparticles are cancer treatment, imaging and diagnostics, antimicrobial agents and hyperthermia therapy⁴. Bimetallic nanoparticles like gold-silver or gold-platinum combinations are used in drug delivery systems for cancer therapy. They provide controlled drug release, enhanced targeting, and reduced side effects by leveraging their unique optical and thermal properties. Bimetallic nanoparticles are employed as contrast agents in bioimaging techniques such as Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) scans. Their enhanced surface plasmon resonance improves imaging quality, aiding in early disease detection. Certain bimetallic nanoparticles, such as silver-copper nanoparticles, show potent antimicrobial effects against bacteria and fungi, addressing the challenge of antimicrobial resistance. Some bimetallic nanoparticles are used in hyperthermia-based cancer treatments, where their ability to generate heat upon laser or magnetic stimulation is exploited to destroy cancer cells selectively⁵. Metal nanoparticles derived from plants, also known as “green-synthesized nanoparticles”, offer an eco-friendly and sustainable approach to nanoparticle production. This method employs plant extracts, which contain natural compounds like phenols, flavonoids, and alkaloids, that act as reducing and stabilizing agents in the synthesis process⁶. Plant-based nanoparticles, such as silver and gold nanoparticles, have shown significant biomedical potential due to their biocompatibility and potent antimicrobial, antioxidant, and anticancer properties⁷. Research continues to explore the potential of various plants in synthesizing nanoparticles, aiming to optimize their biomedical and industrial applications.

Nanocomplex-based controlled release systems are innovative platforms combining nanomaterials with polymers, lipids, or inorganic substances to optimize the delivery and release of active agents. These systems are classified into various types based on their composition and mechanism of action. Polymer-based nanocomplexes, such as polymer–clay and hydrogel nanocomplexes, leverage the flexibility of polymers for tunable release rates, while lipid-based nanocomplexes like liposome-polymer hybrids enhance biocompatibility⁸. Inorganic nanocomplexes, including silica and clay-based systems, offer superior stability and high loading capacity. Stimuli-responsive systems release agents in response to environmental changes, such as pH, temperature, or magnetic fields, making them ideal for targeted therapies⁹. Biodegradable nanocomplexes, often derived from natural polymers, ensure sustainability and safety, while hybrid systems combine organic and inorganic components for multifunctionality. Applications span medicine, agriculture, and food preservation, showcasing their versatility in enhancing efficacy and precision in controlled delivery. Polymer-based nanocomplex release systems offer significant advantages for controlled release applications but face several challenges that hinder their widespread adoption. One major issue is the complexity of manufacturing processes, which often involve multi-step synthesis and precise control, making scalability and cost efficiency difficult¹⁰. Additionally, concerns about biocompatibility and toxicity arise from synthetic polymers and residual materials, which can lead to adverse biological effects¹¹. Stability issues, such as nanoparticle aggregation or polymer matrix deterioration, further limit their storage and functionality. Moreover, many systems suffer from unwanted burst release, where a large portion of the encapsulated agent is rapidly released, reducing efficacy and increasing the risk of side effects¹². Addressing these issues through innovative materials and design strategies is crucial to enhancing the performance and applicability of polymer-based nanocomplexes.

In this study, CuO–ZnO nanoparticles (CuO–ZnO NPs) were green synthesized and used in the composition of nanocomplexes containing chitosan/nanochitosan and P(MMA-co-MAA). Chitosan is extensively used in biomedical applications, including wound healing, drug delivery, and tissue engineering, as it promotes cell growth and has antimicrobial properties¹³. Nanochitosan, a nanometric form of chitosan, exhibits enhanced surface area, solubility, and reactivity compared to bulk chitosan, making it even more effective in drug delivery, especially for controlled release systems. Its small particle size also allows for greater cell uptake and bioavailability, making nanochitosan a preferred material for gene and cancer therapies¹⁴. Poly(methyl methacrylate-co-methacrylic acid) (P(MMA-co-MAA) is a versatile polymer widely studied for drug delivery applications due to its ability to exhibit pH-responsive behavior. In particular, P(MMA-co-MAA) can undergo a significant change in its properties depending on the pH of the environment¹⁵. Under acidic conditions (typical of some diseased tissues or endosomal compartments), the polymer becomes more hydrophilic, leading to an increased rate of drug release. This pH-sensitive behavior allows for controlled drug delivery, where the drug release rate can be fine-tuned based on the surrounding pH, ensuring that the drug is released at the desired location within the body, such as in tumors or inflamed tissues. Furthermore, studies have demonstrated that P(MMA-co-MAA) copolymers can be used in various formulations like hydrogels and nanoparticles for sustained and localized release¹⁶. The aim of this study is to characterize obtained nanocomplexes and determine their potential use in controlled drug release and antimicrobial therapy. Firstly, CuO–ZnO NPs were green synthesized using Salvia officinalis, Cotoneaster horizontalis, Laurus nobilis grown in Turkey. The nanocomplexes (NCs) were synthesized using nanochitosan/chitosan and P(MMA-co-MAA) and bimetal nanoparticles and characterized. Antibacterial and antifungal efficiency and antimicrobial susceptibility tests (Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC)) were applied for nanocomplexes and nanoparticles. BSA and human insulin was used to build the controlled drug release nanocomplexes. The study also reports detailed molecular docking analyses of the obtained CuO–ZnO NPs and NCs. Molecular docking simulations were conducted to provide mechanistic insights into the obtained nanocomplexes. The binding mode of the polymer-stabilized nanocomplex with bovine serum albumin (BSA) and insulin was simulated in order to enlighten the binding mechanism of the synthesized CuO–ZnO hybrid nanocomplex to a drug model by molecular docking analysis. The study recommends a new drug release system including CuO–ZnO NPs with polymers that provide modified release rate. Therefore, the novelty of this study is the development of a green, polymer-stabilized, multifunctional CuO–ZnO hybrid nanocomplex that simultaneously offers enhanced antimicrobial activity, improved structural stability, and sustained protein delivery capability beyond previously reported nanocomposite systems.

Materials and methods

All chemicals and reagents used in this study were of Analytical Grade. Copper(II) chloride dihydrate (CuCl₂·2H₂O, ≥ 99%, AFG Bioscience), zinc acetate dihydrate (Zn(CH₃COO)₂·2H₂O, ≥ 99%, Merck), low-molecular-weight chitosann (50.000–190.000 Da, 75–85% deacetylated, Sigma–Aldrich), sodium tripolyphosphate (TPP, ≥ 85%, Sigma–Aldrich), and bovine serum albumin (BSA, ≥ 98%, Sigma–Aldrich) were used as received. Poly(methyl methacrylate-co-methacrylic acid) [P(MMA-co-MAA)] was obtained from Sigma–Aldrich and prepared as a 1% (w/v) solution in a 1:1 ethanol–water mixture. Human insulin (pharmaceutical grade, Novo Nordisk or equivalent supplier) was used as the model peptide drug. Ethanol (Merck), acetic acid (Sigma–Aldrich), hydrochloric acid (Sigma–Aldrich), and sodium hydroxide (Sigma–Aldrich) were used for solution preparation and pH adjustment. All aqueous solutions were prepared using distilled water.

Biosynthesis of CuO–ZnO NPs

Three plant species of natural Turkish flora were used in the study. Laurus nobilis, Cotoneaster horizontalis and Salvia officinalis were purchased from the Medicinal Plants Garden (Istanbul, Türkiye). Voucher specimens of all plants are preserved in the Department of Bioengineering, Yildiz Technical University. Laurus nobilis leaves, Cotoneaster horizontalis fruits and Salvia officinalis leaves were dried at 80 °C for 48 h. The dried plants were grounded and stirred at 70 °C for 45 min by adding dH₂O. They were then filtered and kept at + 4 °C until use. To prepare bimetallic nanoparticles, 10 mL of plant extract was mixed with 5 mL of 0.1 M CuCl₂ and 5 mL of 0.1 M Zn(C₂H₃O₂)₂ solution. The mixture was stirred overnight in a water bath at 60 °C and 125 rpm. The resulting solution was centrifuged at 5000 rpm for 10 min and the precipitate was washed three times with distilled water. The precipitate was oven dried at 60 °C for 48 h. The same CuO–ZnO NP synthesis procedure was applied for each plant species.

Synthesis of CuO–ZnO NCs

A 1% (w/v) P(MMA-co-MAA) solution was prepared in a 1:1 ethanol–water mixture at room temperature (25 °C) under magnetic stirring. Low-molecular-weight chitosan was dissolved in 1% acetic acid to obtain a 0.3% (w/v) solution and then diluted to 0.01% (w/v) with distilled water. A 1% (w/v) TPP solution was prepared in distilled water and added dropwise (1 mL) to 25 mL of the 0.3% chitosan solution under stirring to form nanochitosan. The mixture was stirred for 20 min and sonicated for 30 min, then diluted to 0.01% (w/v). A 0.1% (w/v) BSA solution was prepared in distilled water and its pH was adjusted to 7.0 prior to incorporation into the nanocomplex formulations. 1 mL of 0.1 mg/mL chitosan solution was mixed with 1 mL of 10 mg/mL P(MMA-co-MAA) while stirring at 120 rpm and 25 °C. The mixture was stirred for 20 min prior to the addition of 1 mg of CuO–ZnO NP solution. The final solution underwent ultrasonication for 30 min at 25 °C, followed by continuous stirring at 120 rpm overnight. The resulting precipitate was separated by centrifugation at 5000 rpm for 20 min, washed and dried at 30 °C for 6 days. For comparison, an identical process was repeated using a 0.01% nanochitosan solution.

Characterization of CuO–ZnO NPs and NCs

To obtain detailed information on the structure, morphology, and size distribution of the CuO–ZnO nanoparticles and nanocomplexes, several analytical techniques were applied, including Fourier Transform Infrared Spectroscopy (FT-IR), Scanning Electron Microscopy (SEM), X-Ray Diffraction (XRD), Zeta-Sizer, and Dynamic Light Scattering (DLS). FT-IR spectra were obtained using a Thermo Fisher Scientific Nicolet iS10 in the range of 400–4000 cm⁻¹. Samples were mixed with KBr and pressed into pellets prior to analysis. For SEM analysis, a drop of the sample suspension was deposited onto a suitable substrate, dried, and analyzed to evaluate morphology and particle distribution using Hitachi S5500. XRD measurements were performed using PhillipsPW1820 x-ray diffractometer with operating CuKα radiation over the 2θ range of 10–80° to determine the structural characteristics and phase behavior of the nanoparticles and nanocomplexes. The average particle size and polydispersity index (PDI) of the CuO–ZnO nanoparticles and nanocomplexes were determined by DLS analysis at room temperature and neutral pH using a Zetasier Malvern Nano ZS. Measurements were performed in triplicate and the results were reported as mean values.

Molecular docking analyses

Preparation and optimization of the covalently bonded chitosan-P(MMA-co-MAA) copolymer structure

The three-dimensional structure of the Chitosan-P(MMA-co-MAA) copolymer was constructed through an integrated workflow combining cheminformatics-based modeling and manual structural (by using Chemdraw Ultra 12 (Advanced Chemistry Development, Inc)) refinement¹⁷. Initially, the simplified molecular fragments of chitosan (represented by glucosamine), methyl methacrylate (MMA), and methacrylic acid (MAA) were generated using RDKit (v2023.03.1) in Python, based on their respective SMILES (Simplified Molecular Input Line Entry System) notations^18,19. Each monomer was converted into a 3D conformer and concatenated into a preliminary linear arrangement to serve as a structural scaffold for subsequent modeling steps. The generated structure was first visualized in Discovery Studio Visualizer and then imported into UCSF Chimera (version 1.19) for detailed structural editing and covalent bonding. Within Chimera, the individual fragments were manually oriented using the Actions > Atoms/Bonds > Move tool to position reactive groups (e.g., hydroxyl and carboxylic termini) in close spatial proximity. Covalent bonds were then explicitly constructed using the Build Structure > Add Bond function by selecting appropriate atom pairs (e.g., the hydroxyl oxygen of the chitosan backbone and the carboxylic carbon of MAA), and assigning the bond type as single. Following bond creation, hydrogen atoms were added via the Structure Editing > AddH function to satisfy valency requirements. The entire structure was then subjected to geometric optimization using the Clean Geometry tool, which resolved unfavorable bond angles and steric clashes introduced during manual manipulation. To ensure a physically realistic conformation, energy minimization was carried out using Chimera’s Minimize Structure module, applying default parameters (100 steps of steepest descent followed by 10 steps of conjugate gradient) under a general molecular mechanics force field. This process refined the geometry and stabilized the polymer architecture. The final covalently bonded and minimized structure was saved in both .mol2 and .pdb formats, rendering it suitable for use in subsequent molecular docking studies as a ligand candidate.

In silico design of CuO–ZnO alloy nanoparticle model

To generate a docking-compatible and dimensionally realistic model of the CuO–ZnO alloy nanoparticle, a revised in silico simulation was performed using Python (v3.10) with NumPy and Matplotlib libraries. In order to prevent steric clashes and computational overloading during docking, the number of atoms was reduced from 1000 to 300, and the simulation box size was adjusted to 2.5 × 2.5 × 2.5 nm. This downsized nanoparticle preserved the alloyed composition, maintaining a 60:40 molar ratio of CuO to ZnO. The atomic coordinates were randomly distributed using numpy.random.rand() within the defined cubic volume, and atom types were assigned based on probability using numpy.random.choice(). To ensure compatibility with docking software, a custom Python script was used to export the structure in PDB format, labeling atoms as “CU” and “ZN” in accordance with PDB standards (Supplementary Information 1). This smaller nanoparticle model was found to be fully compatible with AutoDock Vina docking simulations without exceeding atom limits. The finalized file was saved as .pdb and used as a component in the hybrid nanocomplex for molecular docking studies.

Preparation of the bovine serum albumin (BSA) receptor structure

The crystal structure of BSA was retrieved from the Protein Data Bank (PDB ID: 4F5S), which provides the high-resolution X-ray crystallographic structure of the protein. Prior to molecular docking, the protein structure was preprocessed and optimized using UCSF Chimera (version 1.19). Initially, all non-essential components, including co-crystallized ligands, water molecules, and heteroatoms, were removed to eliminate steric interference during docking. Hydrogen atoms were then added to polar residues to ensure proper protonation states at physiological pH (7.4), and Gasteiger partial charges were assigned automatically using Chimera’s AddH and Add Charge tools. Subsequently, the protein structure was subjected to energy minimization using the Minimize Structure module under the Structure Editing tools^20,21. The minimization protocol included 100 steps of steepest descent followed by 10 steps of conjugate gradient, applying a molecular mechanics force field to relax strained bonds and optimize overall geometry without altering the native tertiary structure. Finally, the minimized and refined receptor structure was saved in both .pdb and .mol2 formats, preparing it for subsequent molecular docking simulations with the Chitosan-P(MMA-co-MAA) copolymer ligand.

Docking simulations

All docking simulations were initially performed using blind docking, in which the grid box was set to encompass the entire receptor structure without prior knowledge of the binding site. Subsequently, docking was refined toward the identified potential binding regions to improve pose prediction accuracy.

(i) Interaction Between Chitosan-P(MMA-co-MAA) and Cu–Zn Alloy Nanoparticle Complex

To evaluate the structural compatibility and interaction potential between the Cu–Zn alloy nanoparticle and the chitosan-based copolymer, an initial docking simulation was conducted. The CuO–ZnO nanoparticle structure, generated via a Python-based in silico approach (as described in Sect.  2.6.2), was loaded as the ligand, and the Chitosan-P(MMA-co-MAA) copolymer served as the receptor. The copolymer consisted of chitosan, methyl methacrylate (MMA), and methacrylic acid (MAA), and their molecular structures were constructed from the following SMILES notations: Chitosan (Glucosamine monomer): NC(C1C(O)C(O)C(O)C(O)O1)CO, MMA (Methyl Methacrylate): CC(=C)C(=O)OC and MAA (Methacrylic Acid): CC(=C)C(=O)O. These structures were converted to 3D via the CACTUS online translator (https://cactus.nci.nih.gov/translate/) and optimized using UCSF Chimera²². The docking was carried out in AutoDock Vina, with the grid centered on the polymeric receptor to capture all potential interaction sites. The resulting complex was energy minimized to be used as a hybrid ligand in subsequent docking protocols.

(ii) Interaction Between Bovine Serum Albumin (BSA) and Chitosan-P(MMA-co-MAA)

In the second docking experiment, the minimized structure of BSA (PDB ID: 4F5S) was used as the receptor, and Chitosan-P(MMA-co-MAA) served as the ligand^21,23. Protein preprocessing was carried out in UCSF Chimera (v1.19) by removing all non-standard residues, heteroatoms, and water molecules. Hydrogen atoms were added, and Gasteiger partial charges were assigned. The docking grid was defined to encompass the entire BSA structure to allow complete binding site accessibility^24,25. Binding affinities were recorded in kcal/mol, and the best-ranked poses were selected for interaction analysis.

(iii) Interaction Between Chitosan-P(MMA-co-MAA) Cu–Zn alloy nanoparticle Coplex and Bovine Serum Albumin (BSA)

In the third simulation, the hybrid structure consisting of Chitosan-P(MMA-co-MAA) bound to Cu–Zn alloy nanoparticles was used as the ligand. The same BSA structure and grid parameters from the previous docking run were reused. This simulation aimed to assess how incorporation of the metallic phase influences protein–polymer interactions. Comparative analysis of binding affinities was used to evaluate whether the presence of the inorganic component enhances or weakens binding to protein surfaces.

(iv) Interaction between ınsulin and chitosan-based carrier systems

In the final set of docking experiments, human insulin (PDB ID: 1ZNJ) was used as the ligand to evaluate its encapsulation affinity toward different chitosan-based nanocomplexes. The insulin structure was preprocessed in UCSF Chimera by removing heteroatoms and water molecules, followed by the addition of hydrogen atoms and assignment of AMBER ff14SB charges²². Nanochitosan was modeled as a trimeric glucosamine unit derived from the chitosan SMILES notation to reflect its reduced molecular weight and compact structure. The 3D structure was generated via RDKit, energy-minimized in UCSF Chimera using 100 steepest descent and 10 conjugate gradient steps, and converted to PDBQT format using AutoDock Tools prior to docking. Three receptor systems were tested: (1) chitosan, (2) nanochitosan, and (3) nanochitosan blended with MMA and MAA. Each receptor was energy minimized and prepared from SMILES-based monomer units, as previously listed. Docking simulations were carried out in AutoDock Vina using grid boxes centered on the nanocarrier structures. Binding energies and binding poses were compared to assess the influence of nanoparticle size and copolymer composition on insulin–carrier interaction dynamics.

(v) Docking simulations with antimicrobial target proteins

To investigate the antimicrobial interaction potential of the designed CuO–ZnO nanoparticle and Chitosan-P(MMA-co-MAA), molecular docking simulations were conducted with four target proteins selected for their roles in bacterial and fungal pathogenicity: CYP51 from Candida albicans (PDB ID: 5V5Z), Sortase A (PDB ID: 1T2P), TolC channel protein (PDB ID: 2OMF), and Protein A (PDB ID: 1BDC). The 3D crystallographic structures of these proteins were obtained from the Protein Data Bank (RCSB PDB) and were preprocessed in UCSF Chimera as previously described in Sect. 2.4.3. Briefly, water molecules, co-crystallized ligands, and heteroatoms were removed, followed by hydrogen addition and energy minimization via the Minimize Structure module using 100 steepest descent and 10 conjugate gradient steps. After minimization, all proteins were converted to PDBQT format using AutoDock Tools, ensuring compatibility with AutoDock Vina 1.2.3²⁶. The Chitosan-P(MMA-co-MAA and CuO–ZnO nanoparticle structures were also prepared in PDBQT format as ligands. Grid box center coordinates were defined individually for each protein based on either the native ligand positions or literature-defined active site residues. Specifically, for crystal structures with co-crystallized ligands, the original ligand coordinates were used as grid centers: C8E for 2OMF (X = –15.4, Y = 43.2, Z = 35.4) and 1YN near the heme site for 5V5Z (X = –37.5, Y = –17.5, Z = 26.2). For proteins without co-crystallized ligands, key functional residues were used to determine active site centers: His120, Cys184, and Arg197 for Sortase A (1T2P) (X = –40.0, Y = –19.4, Z = –1.8), and the IgG-binding surface for Protein A (1BDC) (X = –0.6, Y = –3.2, Z = 1.7). A grid box of 50 × 35 × 35 Å was used for the ketoconazole derivative, while a larger 60 × 60 × 60 Å box was chosen for the CuO–ZnO nanoparticle due to its higher atom count and broader interaction surface. Despite the nanoparticle’s small physical size (~ 2.5 nm), a larger box was necessary to capture its extensive multivalent interactions on the protein surface. The exhaustiveness parameter was set to 32, and 20 binding modes were generated for each docking run. Molecular docking poses and molecular visualizations were generated using UCSF Chimera v1.19 (University of California, San Francisco, CA, USA; https://www.cgl.ucsf.edu/chimera/). Schematic illustrations were initially designed using BioRender (https://www.biorender.com) and subsequently edited and finalized using Microsoft PowerPoint (Microsoft Corporation). Final figure assembly, labeling, and annotation were performed using Microsoft PowerPoint.

Antimicrobial activity analyses

Antimicrobial activity of CuO–ZnO NPs and NCs was measured by well diffusion using Staphylococcus aureus, Escherichia coli and Aspergillus niger. Cultures were obtained in accordance with the 0.5 McFarland’s turbidity standard (10⁸ CFU/mL). 20 µL of bacterial or fungal suspension was spread on agar plates and allowed to dry, followed by the application of 20 µL of nanoparticle and nanocomplex solution (1 mg/mL) to the wells. The positive control was Streptomycin for bacterial samples and Amphotericin B for fungal samples, while distilled water (dH2O) served as the negative control. Incubation temperature was 37 °C and the inhibition zones were measured after 24 h of incubation. E. coli and S. aureus were also employed in MIC and MBC tests. CuO–ZnO nanoparticles and nanocomplexes were diluted from 39 × 10^–4 to 2 mg/mL to assess in vitro sensitivity using the broth dilution method. Incubation parameters were 24 h at 37 °C in nutrient broth. Blank broth served as the negative control, while inoculated broth was the positive control. The lowest concentration that inhibited visible bacterial growth was expressed as the MIC value. Samples taken from MIC tubes without turbidity were cultured in the medium for MBC tests. The lowest concentration that killed bacteria was expressed as the MBC value. All experiments were performed in triplicate.

In vitro release analyses

1 mL of 0.1 mg/mL chitosan/nanochitosan solution and 1 mL of 10 mg/mL P(MMA-co-MAA) were mixed for release testing, while stirring at 120 rpm at 25 °C for 20 min. Then, 1 mL of 0.1 mg/mL BSA (pH 7.0) was added. The mixture was further agitated at 120 rpm for 20 min before incorporating 1 mg of CuO–ZnO NPs. The final mixture underwent ultrasonication for 30 min at 25 °C and was then stirred overnight at 120 rpm. The precipitate was separated by centrifugation at 5000 rpm for 20 min, and the filtrate was collected. The precipitate was washed and dried at 35 °C. 10 mg BSA-loaded NCs were suspended in 1 mL of PBS solution at pH 7.4 including 0.4 mg sodium azide. The BSA-loaded NCs were incubated at 100 rpm and 37 °C. Over a continuous 7-day period, the release medium (PBS solutions) was analyzed with UV–Vis spectrophotometer at a wavelength of 280 nm. As a control, BSA-unloaded nanoparticles underwent the same process and were employed as blank for UV–Vis measurements. To quantify the released BSA, a calibration curve was constructed with the help of free BSA solutions. At 280 nm, various amounts of stock BSA produced in PBS were measured in order to create the calibration curve. The filtrate from the first centrifugation was used to calculate the quantity of unloaded BSA.

In addition to the BSA release experiments, insulin loading and release studies were conducted to further assess the drug delivery potential of the polymer-stabilized nanocomplexes. Human insulin (100 IU/mL, NovoMix®, Novo Nordisk) was loaded into the polymeric nanocomplex system following the same protocol used for BSA. Insulin-loaded nanocomplexes were isolated by centrifugation and used for in vitro release studies in PBS (pH 7.4) at 37 °C under continuous shaking (100 rpm). The released insulin was quantified over 7 days using UV–Vis spectroscopy at 270 nm. Insulin-unloaded samples were used as blanks, and insulin loading capacity (LC%) and cumulative release were calculated using calibration curves prepared with free insulin solutions.

The loading capacity (LC%) was determined using the following formula:

Results

Synthesis of CuO–ZnO nanoparticles and nanocomplexes

In this study, three CuO–ZnO bimetal nanoparticles were synthesized using Cotoneaster fruits, Sage, and Laurel extracts. Then, 6 types of polymer-stabilized nanocomplexes were synthesized from chitosan/nanochitosan, P(MMA-co-MAA) and CuO–ZnO bimetal nanoparticles (Table 1).

Table 1.

Biosynthesized CuO–ZnO nanoparticles (NP) and nanocomplexes (NC).

NP1	CuO–ZnO (Cotoneaster horizontalis)
NP2	CuO–ZnO (Salvia officinalis)
NP3	CuO–ZnO (Laurus nobilis)
NC1	NP1 + Chitosan + P(MMA-co-MAA)
NC2	NP1 + Nanochitosan + P(MMA-co-MAA)
NC3	NP2 + Chitosan + P(MMA-co-MAA)
NC4	NP2 + Nanochitosan + P(MMA-co-MAA)
NC5	NP3 + Chitosan + P(MMA-co-MAA)
NC6	NP3 + Nanochitosan + P(MMA-co-MAA)

Characterization of CuO–ZnO NPs and NCs

Figure 1 shows the SEM micrographs of CuO–ZnO nanoparticles (NP1, NP2, NP3) and their corresponding polymer-stabilized nanocomplexes (NC1–NC6). NP1 exhibited moderately aggregated granular structures, while NP2 displayed smaller and more uniformly dispersed particles. NP3 showed a compact and clustered morphology. Corresponding composites NC1 and NC2 derived from NP1 displayed porous but uneven surfaces. NC3 and NC4, derived from NP2, exhibited relatively compact and homogeneous structures. In contrast, NC5 and NC6 based on NP3 showed larger pores and more irregular morphology. The average particle sizes obtained from SEM were approximately 138 nm for NP1, 62 nm for NP2, and 63 nm for NP3. Among the nanocomplexes, nanochitosan-based samples (NC2, NC4, NC6) consistently appeared denser and more compact compared to their chitosan-based counterparts (NC1, NC3, NC5), indicating differences in matrix organization and dispersion efficiency (Fig. 1).

Fig. 1.

SEM images of the CuO–ZnO NPs (NP 1-3) and NCs (NC 1-6).

The Z-average hydrodynamic diameter (Z-Ave) and polydispersity index (PDI) values of CuO–ZnO nanoparticles (NP1–NP3) and polymer-stabilized nanocomplexes (NC1–NC6) are shown in Fig. 2. Among the nanoparticles, NP1 exhibited the smallest Z-Ave (189.6 nm), followed by NP2 (205.4 nm) and NP3 (207.2 nm). The corresponding nanocomplexes showed increased Z-Ave values, with NC1 displaying the highest value (312.4 nm). Nanochitosan-based samples (NC2, NC4, NC6) showed lower Z-Ave values compared to chitosan-based samples (NC1, NC3, NC5). NP2 exhibited the lowest PDI (0.167), while NP3 showed the highest among nanoparticles (0.231). Among the nanocomplexes, NC3 had the highest PDI (0.301), whereas NC6 exhibited the lowest value (0.197). Overall, nanochitosan-containing samples demonstrated lower PDI values than their chitosan counterparts. The Z-Ave values of nanoparticles were higher than SEM-derived sizes, reflecting hydration effects in aqueous media. All nanoparticle samples exhibited PDI values below 0.3, indicating moderate size uniformity.

Fig. 2.

PDI and Z-average particle size (Z-Ave) values of CuO–ZnO nanoparticles (NPs) and polymer-stabilized nanocomplexes (NCs). Each data point represents a measured value for a specific sample (NP1–NP3 and NC1–NC6). NPs are indicated in blue, and NCs in red. Data labels above each point indicate the exact measured values.

Figure 3 presents the FT-IR spectra of the plant extracts, green-synthesized CuO–ZnO nanoparticles (NP1–NP3), and polymer-stabilized nanocomplexes (NC1–NC6) together with their individual components. The spectra of the plant extracts and CuO–ZnO nanoparticles exhibited marked differences, indicating the structural and compositional changes that occurred during nanoparticle synthesis. The plant extracts showed characteristic absorption bands corresponding to O–H stretching (3200–3500 cm⁻¹), C–H stretching (2800–3000 cm⁻¹), and C=O stretching (around 1700 cm⁻¹), which are associated with phytochemical constituents such as phenolics, flavonoids, terpenoids, polyphenols, and organic acids. These compounds are believed to play a crucial role in the green synthesis process by acting as reducing agents for Cu²⁺ and Zn²⁺ ions and as natural capping agents that stabilize the resulting nanoparticles. Following nanoparticle formation, the FT-IR spectra of NP1–NP3 showed a reduction or disappearance of the characteristic plant-derived bands together with the appearance of new absorption peaks in the 500–600 cm⁻¹ region, corresponding to metal–oxygen vibrations. These bands confirm the formation of CuO and ZnO structures. The observed spectral changes indicate that the phytochemicals in the extracts participated directly in the synthesis process and remained associated with the nanoparticle surface, thereby contributing to nanoparticle stabilization. The FT-IR spectra of the nanocomplexes (NC1–NC6) contained the characteristic bands of both chitosan/nanochitosan and P(MMA-co-MAA). Specifically, chitosan and nanochitosan contributed broad O–H and N–H stretching bands at 3200–3500 cm⁻¹ and amide C=O vibrations near 1650 cm⁻¹, whereas P(MMA-co-MAA) exhibited characteristic C=O stretching at 1730 cm⁻¹ and C–O–C vibrations between 1151 and 1250 cm⁻¹. Compared with the spectra of the individual components, the nanocomplexes showed noticeable shifts and intensity changes in the O–H, N–H, and C=O bands, indicating hydrogen bonding, coordination interactions, and the successful incorporation of CuO–ZnO nanoparticles into the polymeric matrix. In addition, the nanochitosan-containing formulations (NC2, NC4, and NC6) displayed sharper O–H and C=O bands than the corresponding chitosan-based systems, suggesting that the smaller particle size and altered structure of nanochitosan influenced the molecular interactions within the composite. Overall, these FT-IR findings confirm the formation of stable hybrid nanocomplexes and demonstrate strong interfacial interactions among CuO–ZnO nanoparticles, chitosan/nanochitosan, and P(MMA-co-MAA).

Fig. 3.

FT-IR spectra of plant extracts, CuO–ZnO NPs, polymers and nanocomplexes (NP 1-3 and NC 1-6).

The XRD patterns of CuO–ZnO nanoparticles (NP1–NP3) and polymer-stabilized hybrid nanocomplexes (NC1–NC6) are presented in Fig. 4. All samples exhibited broad diffraction halos without sharp Bragg peaks, indicating predominantly amorphous structures. The nanoparticle samples (NP1–NP3) showed diffuse patterns with gradually decreasing intensities across the scanned 2θ range, suggesting the absence of long-range crystalline order. Similarly, the nanocomplexes (NC1–NC6) were characterized by broad halos associated with chitosan, nanochitosan, and P(MMA-co-MAA) matrices. Nanochitosan-based samples (NC2, NC4, NC6) displayed slightly broader and more uniform profiles compared to chitosan-based samples (NC1, NC3, NC5).

Fig. 4.

Comparative XRD profiles of CuO–ZnO nanoparticles and CuO–ZnO/chitosan/nanochitosan–P(MMA-co-MAA) nanocomplexes, highlighting broad diffraction halos indicative of amorphous structures.

Molecular docking analyses

The proposed interaction mechanism between the CuO–ZnO nanoparticles and the chitosan/nanochitosan–P(MMA-co-MAA) polymeric matrix is schematically illustrated in Fig. 5.

Fig. 5.

Schematic illustration of the formation and interaction mechanism of polymer-stabilized CuO–ZnO nanocomplexes, highlighting nanoparticle–polymer coordination, core–shell structure formation, and multivalent interactions with microbial proteins. The schematic was initially designed using BioRender (https://www.biorender.com) and subsequently edited and finalized using Microsoft PowerPoint.

The molecular docking results of the Cu–Zn alloy-modified chitosan-P(MMA-co-MAA) system and its interaction with bovine serum albumin (BSA) are obtained (Fig. 6). The Cu–Zn alloy nanoparticle was successfully integrated onto the surface of the chitosan-based copolymer, forming a stable hybrid structure for docking analysis. The alloy-modified polymer complex exhibited a binding affinity of –10.1 kcal/mol with an RMSD value of 0.889 Å, indicating a stable binding conformation. In addition, the chitosan-P(MMA-co-MAA) copolymer showed a binding affinity of –7.01 kcal/mol toward BSA (Table 2). The interaction involved hydrogen bonding with Arg194, Ser442, and Asp450 residues, while Lys439 contributed to steric effects. The docking poses demonstrate that the polymer and alloy-integrated structures bind within accessible surface regions of BSA, confirming effective molecular recognition and stable protein–carrier interactions.

Fig. 6.

Molecular docking representation of the Cu–Zn alloy-modified chitosan P(MMA-co-MAA) composite and its interaction with bovine serum albumin (PDB ID: 4F5S). The upper panel shows the Cu–Zn alloy nanoparticle anchored to the chitosan-based copolymer surface. The lower panel illustrates the binding pose of the polymer complex at the BSA surface, with a zoomed-in view highlighting interacting residues (Ser442, Asp450, Lys439). Visualization generated using UCSF Chimera v1.19 (https://www.cgl.ucsf.edu/chimera/). Labels and figure assembly were completed using Microsoft PowerPoint.

Table 2.

Summary of molecular docking results of CuO–ZnO nanoparticles and polymer-stabilized systems.

Complexes	Target Protein	Binding Affinity (kcal/mol)	RMSD (Å)
Cu–Zn NP + Polymer	BSA	– 10.10	0.89
Polymer only	BSA	– 7.01	1.02
CuO–ZnO NP	Sortase A	– 70.06	0.95
CuO–ZnO NP	TolC	– 62.43	1.08
CuO–ZnO NC	CYP51	– 73.48	0.87
CuO–ZnO NP	Protein A	– 40.80	1.10
Chitosan	Insulin	– 6.20	1.05
Nanochitosan	Insulin	– 6.98	0.92
Chitosan–P(MMA-co-MAA)	Insulin	– 7.56	0.47

Molecular docking analysis revealed that the CuO–ZnO nanoparticle exhibits significantly higher binding affinities to bacterial and fungal target proteins compared to the chitosan-P(MMA-co-MAA), suggesting enhanced multivalent interactions with the protein surfaces (Figs. 7 and 8). Specifically, the nanoparticle showed strong interaction with Sortase A (PDB: 1T2P) and the TolC outer membrane channel (PDB: 2OMF), yielding docking scores of –70.06 and –62.43 kcal/mol, respectively, which are markedly higher than those of the reference ligand (–6.02 and –5.93 kcal/mol) (Table 2). These findings imply that the nanoparticle establishes broader and more stable contact with functionally important domains, potentially disrupting bacterial adhesion and efflux mechanisms (Fig. 7). Furthermore, the chitosan-P(MMA-co-MAA)-stabilized CuO–ZnO nanocomplex exhibited a remarkably high binding affinity to CYP51 (PDB: 5V5Z), a critical fungal enzyme, with a docking score of –73.48 kcal/mol, greatly surpassing the ketoconazole derivative (–6.745 kcal/mol), likely due to strong interactions at the heme-binding region. Additionally, the nanoparticle demonstrated effective binding to Protein A (PDB: 1BDC), a bacterial immune evasion protein, with a docking score of –40.80 kcal/mol, compared to –5.21 kcal/mol for the ligand (Fig. 8).

Fig. 7.

Molecular docking representations of CuO–ZnO nanoparticles and chitosan-P(MMA-co-MAA) with Sortase A (PDB: 1T2P) and TolC outer membrane protein (PDB: 2OMF), both bacterial virulence-related proteins. Visualization generated using UCSF Chimera v1.19 (https://www.cgl.ucsf.edu/chimera/). Labels and figure assembly were completed using Microsoft PowerPoint.

Fig. 8.

Molecular docking of CuO–ZnO nanoparticles and chitosan-P(MMA-co-MAA) with IgG-binding protein A (PDB: 1BDC) and fungal CYP51 (PDB: 5V5Z). Visualization generated using UCSF Chimera v1.19 (https://www.cgl.ucsf.edu/chimera/). Labels and figure assembly were completed using Microsoft PowerPoint.

Molecular docking results demonstrated that insulin interacted differentially with chitosan-based nanocarriers depending on structural modifications (Fig. 9). The native chitosan system yielded a moderate binding affinity (–6.20 kcal/mol), engaging His10 and Arg9 residues within the insulin molecule. In contrast, nanochitosan displayed a deeper encapsulation pattern with improved binding energy (–6.98 kcal/mol), likely due to its higher surface area and denser network, which enhances hydrogen bonding and van der Waals interactions. Notably, the chitosan–P(MMA-co-MAA) copolymer exhibited the most favorable binding (–7.56 kcal/mol) with the lowest RMSD (0.47 Å), indicating a more stable and specific interaction at peripheral residues Glu13 and Ala14 (Table 2).

Fig. 9.

Molecular docking visualization of insulin with chitosan-based nanocarriers. Visualization generated using UCSF Chimera v1.19 (https://www.cgl.ucsf.edu/chimera/). Labels and figure assembly were completed using Microsoft PowerPoint.

Antimicrobial activity of the CuO–ZnO NPs and NCs

The antimicrobial activity of CuO–ZnO nanoparticles (NP1–NP3) and polymer-stabilized nanocomplexes (NC1–NC6) against Escherichia coli, Staphylococcus aureus, and Aspergillus niger is shown in Table 3. Among the nanoparticles, NP2 exhibited the highest inhibition zones, measuring 19 mm against E. coli, 21 mm against S. aureus, and 26 mm against A. niger. NP3 showed inhibition zones of 18–23 mm, while NP1 exhibited comparatively lower activity. The polymer-stabilized nanocomplexes showed enhanced antimicrobial performance, with NC4 displaying the highest overall activity (22 mm against E. coli, 24 mm against S. aureus, and 25 mm against A. niger). NC6 also demonstrated strong inhibition, particularly against E. coli (21 mm) and A. niger (25 mm). Nanochitosan-based samples (NC2, NC4, NC6) generally showed higher inhibition zones than their chitosan-based counterparts. The positive control streptomycin exhibited inhibition zones of 21 mm for E. coli and 24 mm for S. aureus, while amphotericin B showed 28 mm against A. niger. Negative controls showed no inhibitory activity.

Table 3.

Agar well diffusion data of the CuO–ZnO NPs and NCs (mm).

	E. coli	S. aureus	A. niger
NP1	15 ± 1	17 ± 1	21 ± 1
NP2	19 ± 2	21 ± 2	26 ± 2
NP3	18 ± 2	20 ± 1	23 ± 1
NC1	17 ± 1	19 ± 1	23 ± 1
NC2	19 ± 1	21 ± 1	24 ± 2
NC3	20 ± 1	22 ± 2	24 ± 1
NC4	22 ± 2	24 ± 2	25 ± 2
NC5	20 ± 2	21 ± 1	24 ± 1
NC6	21 ± 1	21 ± 1	25 ± 2
Negative control (dH2O)	-	-	-
Positive control (Streptomycin)	21 ± 1	24 ± 1	-
Positive control (Amphotericin B)	-	-	28 ± 2

The MIC and MBC values of CuO–ZnO nanoparticles and their polymer-stabilized nanocomplexes against E. coli and S. aureus are obtained (Table 4). Both CuO–ZnO NPs and NCs exhibited identical MIC values of 0.50 mg/mL against E. coli and 0.25 mg/mL against S. aureus. The MBC values were also the same across both formulations, with 0.50 mg/mL for E. coli and 0.25 mg/mL for S. aureus. No variation was observed in antibacterial concentration thresholds between nanoparticles and their polymer-based nanocomplexes.

Table 4.

MIC and MBC analysis results of the CuO–ZnO NPs and NCs.

Sample	MIC (mg/mL) E. coli	MIC (mg/mL) S. aureus	MBC (mg/mL) E. coli	MBC (mg/mL) S. aureus
CuO–ZnO NP	0.50	0.25	0.50	0.25
CuO–ZnO NC	0.50	0.25	0.50	0.25

In vitro drug release of BSA model

Figure 10 presents the in vitro release profiles of BSA from two polymer-stabilized CuO–ZnO hybrid nanocomplexes over a 7-day period at pH 7.4 and 37 °C. Both the chitosan-based and nanochitosan-based systems exhibited a biphasic release pattern, with an initial rapid release phase (0–4 h) followed by a more sustained release phase extending to 7 days. The chitosan formulation achieved a cumulative release of 99% by day 7, whereas the nanochitosan formulation reached 94% release. Throughout the time course, both systems demonstrated comparable release behaviors, with the chitosan formulation showing slightly higher release percentages at most time points.

Fig. 10.

In vitro BSA release profiles of polymer-stabilized CuO–ZnO hybrid nanocomplexes at pH 7.4 and 37 °C. (1) Chitosan/P(MMA-co-MAA)/BSA/CuO–ZnO NPs and (2) Nanochitosan/P(MMA-co-MAA)/BSA/CuO–ZnO NPs.

The in vitro human insulin release profiles from polymer-stabilized CuO–ZnO hybrid nanocomplexes at pH 7.4 and 37 °C over 7 days are obtained (Fig. 11). Both the chitosan-based (Chitosan/P(MMA-co-MAA)/Insulin/CuO–ZnO NPs) and nanochitosan-based (Nanochitosan/P(MMA-co-MAA)/Insulin/CuO–ZnO NPs) systems showed sustained release behavior without a significant initial burst. At 2 h, the cumulative insulin release reached approximately 25% for chitosan-based and 17% for nanochitosan-based nanocomplexes. Over the course of 7 days, cumulative release increased steadily, reaching 68% for the chitosan-based system and 60% for the nanochitosan-based system. Throughout the experiment, the chitosan-based formulation maintained a slightly higher release rate compared to the nanochitosan-based system. The incorporation of P(MMA-co-MAA) significantly influenced the release behavior of both BSA and insulin. Although methyl methacrylate-rich polymers are generally considered relatively hydrophobic, the copolymer employed in the present study contained both hydrophobic methyl methacrylate (MMA) units and hydrophilic, ionizable methacrylic acid (MAA) groups. Consequently, the release behavior was governed by the balance between these two structural components. The MMA segments reduced water penetration and restricted rapid swelling of the polymer matrix, thereby limiting premature diffusion of the loaded biomolecules. In contrast, at pH 7.4 the carboxylic acid groups of the MAA units became partially ionized, increasing matrix hydration and facilitating gradual diffusion-controlled release. As a result, both polymer-stabilized CuO–ZnO nanocomplexes exhibited sustained release profiles over 7 days with only a limited initial burst. The BSA-loaded systems displayed a biphasic pattern consisting of an initial rapid release phase followed by a slower sustained release stage, while the insulin-loaded systems showed a more gradual release throughout the experiment. The chitosan-based formulation exhibited slightly higher cumulative release values than the nanochitosan-based system for both biomolecules. By day 7, cumulative BSA release reached 99% for the chitosan formulation and 94% for the nanochitosan formulation, whereas cumulative insulin release reached 68% and 60%, respectively. The lower release observed in the nanochitosan-containing systems may be attributed to the formation of a denser and more compact polymeric network, leading to stronger intermolecular interactions between nanochitosan, P(MMA-co-MAA), and the loaded biomolecules. Therefore, the principal role of P(MMA-co-MAA) in the present system was to provide a pH-responsive and diffusion-regulating matrix capable of modulating sustained biomolecule release.

Fig. 11.

In vitro human insulin release profiles of polymer-stabilized CuO–ZnO hybrid nanocomplexes at pH 7.4 and 37 °C. (1) Chitosan/P(MMA-co-MAA)/Insulin/CuO–ZnO NPs and (2) Nanochitosan/P(MMA-co-MAA)/Insulin/CuO–ZnO NPs.

Discussion

The integration of green-synthesized bimetallic CuO–ZnO nanoparticles with chitosan-based polymeric matrices represents a promising strategy for the development of multifunctional nanocomplex platforms for biomedical applications^27,28 and the present system follows a similar multivalent assembly pathway leading to a core–shell nanocomplex architecture with enhanced structural integrity and biological functionality (Fig. 5). In this study, a rational workflow combining in silico modeling, physiochemical characterization, and in vitro evaluations was employed to assess structural organization, antimicrobial performance, and drug/protein release dynamics of six CuO–ZnO nanocomplex formulations synthesized using plant extracts and chitosan/nanochitosan–P(MMA-co-MAA) copolymers. SEM, DLS, FT-IR, and XRD analyses collectively confirmed the successful synthesis and structural integration of the nanocomplexes. Notably, the SEM data revealed that particle morphology and dispersion were significantly influenced by the plant extract used in nanoparticle synthesis (Fig. 1), supporting earlier findings that phytochemical composition governs nanoparticle nucleation and stabilization behavior during green synthesis^6,29. NP2, synthesized using Salvia officinalis, yielded the smallest and most homogeneously distributed particles (62 nm), which translated into denser and more uniform nanocomplexes (NC3, NC4). This was further corroborated by Z-Ave and PDI data (Fig. 2), where NC4 exhibited one of the lowest hydrodynamic diameters and size distributions, reflecting enhanced matrix compatibility and particle dispersion.

In the FT-IR analyses, new peaks appeared around 500–600 cm^—1 in all NP spectra, which can be attributed to Cu–O and Zn–O stretching vibrations, confirming the formation of bimetallic CuO–ZnO NPs. NP1 showed peaks at 537 cm⁻¹, NP2 at 584 cm⁻¹, and NP3 at 597 cm⁻¹, indicative of metal–oxygen bonds (Fig. 3). These changes demonstrated that the plant extract biomolecules facilitate the green synthesis of CuO–ZnO NPs and stabilizing their structure. The successful formation of NPs was evident from the appearance of characteristic metal–oxygen peaks and the modifications in the functional group regions. The slightly sharper peaks in nanochitosan-based NCs (NC2, NC4, NC6) may reflect improved structural order and hydrogen bonding potential due to the smaller size and higher surface area of nanochitosan^30–32. XRD patterns of NP1–NP3 and NC1–NC6 exhibited broad diffraction halos without distinct Bragg reflections, indicating predominantly amorphous structures (Fig. 4). This behavior is consistent with previous reports showing that plant extract-mediated synthesis and polymer encapsulation suppress long-range crystallinity due to phytochemical capping and the absence of thermal annealing^33,34. Under crystalline conditions, ZnO typically exhibits the characteristic reflections of a hexagonal wurtzite phase, whereas CuO shows monoclinic diffraction peaks. However, no separate ZnO- or CuO-related reflections were observed in the present samples. Instead, the strong interactions between CuO–ZnO nanoparticles, phytochemicals, and the chitosan/nanochitosan–P(MMA-co-MAA) matrix restricted crystal growth and reduced the coherence length of the inorganic phase. As a result, the XRD profiles became dominated by the amorphous polymeric network, masking the expected crystalline signatures of ZnO and CuO. The slightly broader halos observed for nanochitosan-based systems further suggest a more compact and homogeneous dispersion of the nanoparticles within the matrix.

In silico docking provided critical insights into the molecular interaction capabilities of the CuO–ZnO NCs before experimental evaluation. The docking of the chitosan-P(MMA-co-MAA)/Cu–Zn composite to bovine serum albumin (BSA) showed strong binding (–10.1 kcal/mol) with minimal RMSD variation (0.889 Å), confirming a stable complex (Fig. 6). Binding occurred via hydrogen bonding and van der Waals contacts, consistent with known affinity of polysaccharide-based nanocarriers for serum proteins, enabling prolonged systemic circulation without irreversible binding^35,36. Importantly, the stepwise docking analysis validated that the Cu–Zn alloy nanoparticle has a significantly stronger binding affinity to the copolymer (–10.1 kcal/mol) than the resulting nanocomplex does to BSA (–7.01 kcal/mol). This binding hierarchy suggests a thermodynamically favorable interaction cascade, ensuring that the nanoparticle is first efficiently encapsulated by the polymer, followed by moderate reversible interactions with plasma proteins—an ideal scenario for drug delivery carriers^37,38. Docking studies targeting bacterial (Sortase A, TolC) and fungal (CYP51) virulence proteins demonstrated the potential antimicrobial mechanism of the CuO–ZnO NPs (Figs. 7 and 8). Nanoparticles showed remarkably high docking scores with all targets (up to –73.48 kcal/mol for CYP51), significantly surpassing the reference chitosan-P(MMA-co-MAA) (–6.745 kcal/mol), suggesting extensive surface contact and multivalent interactions (Table 2)^39–42. The unusually high binding affinities (e.g., ~  − 70 kcal/mol) observed in this study should not be interpreted as conventional ligand binding energies^43,44. These values arise from the multivalent interaction nature of nanoparticle systems, where multiple simultaneous contacts between the nanoparticle surface and protein residues contribute cumulatively to the docking score. Such behavior differs fundamentally from small-molecule docking and is known to produce exaggerated binding energies regardless of the docking software used⁴⁵. The multivalent and metallic nature of the nanoparticles likely enables broader and more stable interaction networks on protein surfaces, differing mechanistically from classical ligand–receptor binding, which typically engages smaller and more specific binding pockets^11,38. These predictions were later confirmed in vitro, highlighting the utility of in silico screening for prioritizing candidate formulations and identifying optimal targets.

The molecular docking findings of the present study are consistent with and extend those reported in recent literature on CuO–ZnO-based nanocomposites (Table 2). Ghosh et al. developed a soluble starch-coated CuO–ZnO nanocomposite loaded with chloramphenicol and demonstrated enhanced antibacterial and anti-biofilm efficacy against E. coli and S. aureus, highlighting the advantage of polymer coating in improving the bioactivity and stability of CuO–ZnO systems⁴⁶. Similarly, Dhanalakshmi et al. fabricated multifunctional CuO–ZnO nanocomposites via green synthesis using Ceropegia debilis extract and conducted molecular docking computational investigations confirming favourable interactions with antimicrobial target proteins, alongside strong antimicrobial activity against S. aureus and E. coli⁴⁷. Abada et al. reported that a biogenically synthesised CuO–ZnO nanocomposite produced using spent mushroom substrate exhibited notable antifungal activity against Candida albicans, with an inhibition zone of 33.5 ± 2 mm, and molecular docking against CYP51 confirmed favourable interactions at the active site, supporting the fungal membrane-targeting mechanism of CuO–ZnO systems⁴⁸. In a related study, Pandey et al. demonstrated that mesoporous ZnO–CuO composite nanoparticles interacted with bacterial and fungal virulence proteins including PqsR, RstA, FosA, and Hsp90, with binding affinities ranging from –5.17 to –7.50 kcal/mol for single-metal oxides; crucially, the bimetallic ZnO–CuO composite reduced inhibition constants to the nanomolar range, underscoring the synergistic advantage of the bimetallic composition⁴⁹. Furthermore, El-Sayed et al. reported that Cu-doped ZnO nanoparticles exhibited binding affinities of –5 to –8 kcal/mol against FabH of E. coli and penicillin-binding proteins of S. aureus, with docking interactions mediated primarily through hydrogen bonding and hydrophobic contacts at the active sites⁵⁰. Collectively, these studies confirm that CuO–ZnO-based systems consistently interact with key bacterial and fungal virulence targets, and the substantially higher binding scores obtained in the present study (–62.43 to –73.48 kcal/mol) reflect the extended multivalent interaction surface of the 300-atom nanoparticle model, which enables simultaneous engagement of multiple protein surface residues—a mechanistic advantage over both single-metal oxides and smaller nanoparticle models employed in prior work. Regarding the significant changes observed between bare CuO–ZnO nanoparticles and the chitosan/nanochitosan–P(MMA-co-MAA)–ZnO–CuO composite system, the polymer matrix was found to modulate the interaction profile of the nanoparticle in a target-dependent manner. For bacterial targets, the bare nanoparticle achieved higher docking scores against Sortase A (–70.06 kcal/mol) and TolC (–62.43 kcal/mol), attributed to the unobstructed metal surface enabling direct multivalent contacts with protein residues. In contrast, the polymer-stabilized composite exhibited enhanced binding toward CYP51 (–73.48 kcal/mol), suggesting that the chitosan/P(MMA-co-MAA) framework orients the nanoparticle toward the heme-binding domain and introduces additional hydrogen bonding and hydrophobic contacts at the interface. Furthermore, nanochitosan-containing systems demonstrated lower RMSD values compared to chitosan-based counterparts, indicating improved binding stability, consistent with their denser polymeric network observed in physicochemical characterisation.

In the present study, molecular dynamics (MD) simulations were not performed, and the interaction stability was evaluated based on molecular docking results. While docking is widely used to predict binding affinity and identify potential interaction regions⁵¹, we acknowledge that it provides a static approximation and does not fully capture the dynamic behavior of the system. It is important to note that the application of conventional MD simulations to nanoparticle–protein systems presents significant methodological challenges. Unlike classical small-molecule ligands, metal oxide nanoparticles such as CuO–ZnO do not possess well-defined atomistic topologies or parameterized force fields within standard MD frameworks^52,53. The presence of mixed metal centers (Cu²⁺/Zn²⁺), surface heterogeneity, and multivalent interaction sites makes accurate parametrization highly complex^54,55. In addition, the relatively large size and irregular structure of nanoparticle models substantially increase computational cost, often requiring advanced coarse-grained or reactive force field approaches that are still under active development^56,57. Therefore, in this study, a docking-based approach was adopted as a first-level screening tool to evaluate the interaction potential of the nanocomplex system.

The antimicrobial assays revealed that all CuO–ZnO nanocomplexes displayed significant antibacterial and antifungal activity, with NC4 showing the highest inhibition across all tested organisms (Table 3). This aligns with the in silico predictions, where the highest binding energies were also observed for CuO–ZnO targeting bacterial and fungal proteins. The enhanced activity of nanochitosan-based samples (NC2, NC4, NC6) compared to their chitosan-based analogs is attributed to better nanoparticle dispersion and increased surface area, facilitating closer interaction with microbial membranes^32,58,59. The synergistic antimicrobial effect of the bimetallic CuO–ZnO nanoparticles is attributed to the complementary actions of the two oxides. ZnO primarily enhances Reactive Oxygen Species generation and membrane disruption, whereas CuO promotes Cu²⁺ ion release and intracellular oxidative damage. Their combination increases membrane permeability, oxidative stress, and overall microbial inhibition more effectively than either oxide alone. In addition, smaller and well-dispersed nanoparticles are generally more active because they generate higher levels of ROS and release metal ions more efficiently, while the chitosan-based matrix may further support nanoparticle adhesion to microbial surfaces and contribute additional membrane-disruptive effects. The observed sensitivity of both bacterial strains and Aspergillus niger is therefore consistent with previously reported CuO–ZnO systems⁶⁰. Comparable inhibition zones to positive controls (streptomycinand amphotericin B) reinforce the potential of these green-synthesized nanocomplexes as alternative or adjunct antimicrobial agents^61,62. Interestingly, the MIC and MBC values were identical for both nanoparticles and nanocomplexes (Table 4), suggesting that antimicrobial efficacy is primarily driven by the CuO–ZnO nanoparticles themselves, while the polymer matrix enhances dispersion and stability but does not independently reduce the required concentrations. This is consistent with literature reports emphasizing the role of polymer matrices in modulating release profiles and system stability, rather than directly contributing to bactericidal activity^63,64. For comparison, several other metal oxides have also been extensively investigated in antimicrobial and drug-delivery applications. Ag₂O exhibits strong broad-spectrum antimicrobial activity due to Ag⁺ release, but its use may be limited by cost and cytotoxicity. TiO₂ and MgO are attractive because of their low toxicity and ROS-related antimicrobial effects. In contrast, AuO is highly biocompatible and useful for drug delivery and imaging, but generally has lower intrinsic antimicrobial activity⁶⁵. Therefore, the CuO–ZnO system employed in the present study offers a favorable balance between antimicrobial efficacy, cost, and compatibility with polymer-based controlled release systems.

The in vitro release studies using BSA and insulin as model proteins demonstrated that both chitosan- and nanochitosan-based nanocomplexes achieve sustained release over 7 days (Figs. 10 and 11). The BSA release profile followed a biphasic pattern—an initial burst (0–4 h) followed by gradual release—typical of many chitosan-based systems due to surface desorption followed by diffusion from the polymer matrix^12,66. Notably, nanochitosan-based systems showed slightly slower and more controlled release profiles, likely resulting from denser matrix packing and stronger drug–polymer interactions^67–69. These effects are particularly valuable for proteins like insulin, where a steady release rate is preferable to avoid plasma concentration spikes^70–72. The insulin release study further emphasized that both systems effectively prevented a significant burst effect, with cumulative release reaching 68% (chitosan) and 60% (nanochitosan) by day 7. These results compare favorably with conventional chitosan nanoparticle systems, which often display faster release and incomplete retention^70,72,73. The amorphous nature of the nanocomplexes, confirmed by XRD, likely contributed to the reduced diffusion rates and sustained release^33,74. Moreover, the pH-responsive character of the P(MMA-co-MAA) matrix may have provided additional regulatory control over insulin mobility under physiological pH conditions^15,16. This study presents a green, biocompatible, and functionally tunable nanoplatform for dual antimicrobial and protein delivery applications. Also a comparison of the antimicrobial activity and release behavior of the CuO–ZnO/chitosan/nanochitosan–P(MMA-co-MAA) nanocomplexes with previously reported metal oxide-based nanocomposite systems were shown in Table 5. The sequential workflow—from plant-mediated synthesis and polymer hybridization to computational modeling and in vitro testing—demonstrates the importance of combining in silico methods with experimental validation. Pre-docking to therapeutic and microbial protein targets enabled early assessment of interaction patterns and provided a predictive framework that was later confirmed through empirical data. Furthermore, replacing bulk chitosan with nanochitosan improved nanoparticle dispersion, protein interaction dynamics, and release control, affirming the impact of nanoscale structural tuning on nanocarrier performance^13,67,75,76. Given their significant antibacterial and antifungal activity, sustained release profiles, and stability in biological environments, these CuO–ZnO polymer-stabilized nanocomplexes hold strong potential for future application in drug delivery systems, wound healing materials, or implant coatings.

Table 5.

Comparison of the antimicrobial activity, release behavior, and multifunctional properties of the present CuO–ZnO/chitosan/nanochitosan–P(MMA-co-MAA) nanocomplexes with previously reported metal oxide-based nanocomposite systems.

System	Matrix/composition	Antimicrobial activity	MIC	Release behavior	Main advantage
Present study	CuO–ZnO/chitosan–P(MMA-co-MAA)	15–22 mm against E. coli; 17–24 mm against S. aureus; 21–25 mm against A. niger	0.50 mg/mL (E. coli); 0.25 mg/mL (S. aureus)	Sustained BSA and insulin release over 7 days with low burst effect	Dual antimicrobial and protein delivery functionality
GA@CuO–ZnO nanocomposite	CuO–ZnO/GA	21–24 mm against fungi	31.25 to 62.5 μg/mL	Not evaluated	Good antifungal activity but no biomolecule release capability62
Chitosan derivative nanoparticles	Trimethyl chitosan-based systems	Not evaluated	Not evaluated	74–83% cumulative release, high burst during first 30 min	Rapid release under physiological conditions71
Chitosan–CuO nanocomposite	CuO/chitosan	6–24 mm depending on strain	0.01–0.1%	Not evaluated	Good antibacterial effect but no controlled release function77
ZnO/chitosan nanocomposite	ZnO/chitosan/MWCNT	Strong activity against H. pylori	12.5–25 μg/mL	Not evaluated	Enhanced antibacterial activity, but limited to antibacterial application only78
Ag/ZnO/chitosan nanocomposite	Ag/ZnO/chitosan	Effective against bacteria and fungi	8 μg/mL	Not evaluated	High antimicrobial activity but higher cost and possible cytotoxicity79
Conventional chitosan nanoparticles	Chitosan/TPP nanoparticles	Not evaluated	Not evaluated	44–60% cumulative release, high burst effect (~ 45% within first 6–12 h)	Pronounced burst release and short release duration80
Alginate-coated chitosan microparticles	Chitosan/alginate microparticles	Not evaluated	Not evaluated	50% retained after 48 h, high burst effect (~ 84% within first 0.5 h for uncoated system)	Strong burst release without coating81
PLGA–chitosan composite nanoparticles	PLGA/chitosan nanoparticles	Not evaluated	Not evaluated	78% cumulative release, moderate burst (~ 32% in first 4 h)	Shorter sustained release than present system82

Conclusions

In conclusion, this study successfully developed a biocompatible, plant-mediated CuO–ZnO nanoparticle system stabilized within chitosan/nanochitosan and P(MMA-co-MAA) matrices, exhibiting multifunctionality for biomedical use. The combination of experimental characterization and in silico modeling demonstrated that these nanocomplexes possess favorable structural, antimicrobial, and drug-delivery properties. Despite numerous studies on monometallic CuO or ZnO nanoparticles, limited attention has been paid to green-synthesized bimetallic CuO–ZnO systems simultaneously integrated with chitosan/nanochitosan and P(MMA-co-MAA) matrices. Furthermore, previous studies have generally focused either on antimicrobial activity or on drug-delivery behavior alone. Thus, a significant research gap exists in the development of multifunctional nanocomplexes capable of combining potent antimicrobial activity with sustained biomolecule release. This gap was addressed through the fabrication of green-synthesized CuO–ZnO nanoparticles incorporated into both chitosan- and nanochitosan-based P(MMA-co-MAA) matrices. The principal novelty of the present work lies in the integrated evaluation of antimicrobial activity, molecular docking interactions, and controlled release performance within a single platform, together with a direct comparison between chitosan- and nanochitosan-based systems. In particular, nanochitosan-containing formulations reduced burst release and provided a more sustained release profile, thereby overcoming important limitations associated with conventional chitosan-based nanocomposites. Moreover, the molecular docking results revealed strong binding affinities toward key bacterial and fungal target proteins, further supporting the therapeutic potential of the developed nanocomplexes. Therefore, this green-synthesized hybrid system represents a cost-effective and environmentally friendly nanocarrier platform with promising applicability in antimicrobial therapy and sustained drug/protein delivery. Nevertheless, the present study has certain limitations, including the absence of BET measurements for pore size distribution, EDX/ICP-based elemental quantification, SAED confirmation of crystallinity, and cytotoxicity evaluation on mammalian cells. These aspects should therefore be addressed in future studies to further validate the structural characteristics, elemental distribution, and biocompatibility of the developed nanocomplexes.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Ecem Isiksel: Investigation, Formal analysis.Azade Attar: Writing – original draft, Data curation.Emre Aktas: Computational modeling, Writing – original draft.Melda Altikatoglu Yapaoz: Supervision, Conceptualization.

Funding

This study was financially supported by the Scientific Research Commission of Yildiz Technical University (project no. FBA-2024-6436).

Data availability

All data generated or analyzed during this study are included within the manuscript.

Declarations

Competing interests

The authors declare no competing interests.