Introduction

Targeted drug delivery (TDD) and nanomedicine collaborate to improve the safety and efficacy of pharmaceuticals and, in turn, their therapeutic activity in vivo by precisely regulating the distribution of drugs to the potential target site across a range of biological barriers1. Because nanocarriers are made to overcome obstacles in the body, they can target particular regions and have a therapeutic impact2. Researchers have looked into several strategies to produce the best carriers3. Drug carriers are often created and produced by the intended target conditions, and their performance is subsequently assessed in vitro and in vivo, respectively4. The wide range of biomedical uses of nanocarriers, including immunotherapy, drug delivery, and tissue regeneration, accelerates the research of materials, formulations, and production techniques for nanocarriers5. A medication must remain stable in the bloodstream for it to reach its target location and provide the desired effects. When any drug formulation, especially a nanocarrier system, enters the bloodstream, these proteins form a protein-protein combination with nanoparticles (NPs)6.

Several molecular-level chemical, biological, and physical interactions control the establishment of this complex. As a result, how well nanoparticles interact with protein complexes determines their fate in vivo. Synthetic and natural materials are used to create and deliver NPs7. Chitosan holds a unique position among the natural materials utilized in biomedical applications thus far because of its hydrophilicity, low toxicity, biocompatibility, biodegradability, and structural variety8. Despite its special qualities, chitosan’s uses are restricted because of its subpar mechanical attributes. Nevertheless, it is a viable and stable drug delivery carrier for diagnostics and treatment9.

A number of tactics have been developed to address these problems. For example, because of the material’s high affinity for useful proteins and capacity for self-assembly, free amine and hydroxyl groups have produced a variety of chitosan derivatives with improved solubility10. Chitosan-conjugated elements in target tumors, sometimes referred to as eudragit, also respond to chemical or physical stimuli from the outside and the inside11. Additionally, methacrylate monomers include dimethylamino methacrylate, also called eudragit, which is frequently used in pharmaceutical applications, as well as methacrylic acid and methacrylic acid esters12.

Despite sharing a common structure, these polymers differ in the substituents’ chemical characteristics. Their solubility has been studied in various organic solvents, and their pH-dependent water solubility makes them perfect coatings for drug delivery devices13. Stimuli-responsive polymeric nanocarriers have shown great promise in medication delivery applications due to their increased bioavailability at the target location14. In this instance, the drug is released in response to a variety of particular internal or external stimuli. It is essential to construct the molecular structure in response to stimuli.

Biological activities are based on molecular interactions impacted by macromolecular structures. Therefore, mathematical computations and research are crucial before beginning experimental and pilot studies.

A molecular dynamics (MD) simulator is useful for comprehending macromolecular structure-function interactions. The factors of interest are thoroughly investigated at the atomic level to better understand these techniques. The data on the macromolecule’s dynamic properties are trustworthy enough to create a bioinformatic structural model15. Numerous macromolecules can be represented and analyzed via simulation, which allows for debate and investigation of a range of therapy possibilities. Molecular dynamics can be simulated using various techniques, including all-atom molecular dynamics simulation (AA-MD) and coarse-grained molecular dynamics simulation (CG-MD)16. The AA-MD simulator simulates every atom independently. It is expensive but computationally accurate, and it is used for smaller systems17. Its uses include demonstrating and comprehending the structure of polymers, lipid bilayers, and soluble proteins. This kind of simulation is limited to time scales of tens of nanoseconds due to the computational cost of simulating the interactions between the solute and solvent. CG-MD force fields were developed to get around this restriction. In this instance, a single CG-MD bead represents groupings of atoms. Longer simulations can be done using the CG-MD approach18. AA-MD and CG-MD methodologies could replicate a variety of medications and their delivery methods.

Doxorubicin (DOX), a broad-spectrum anticancer medication, has demonstrated therapeutic effects on various solid tumor types19. DOX targets several molecular locations and impairs macromolecular formation by interacting with DNA, resulting in various cytotoxic consequences. DOX is highly harmful to both normal and malignant cells because of its weak tumor selectivity20. In particular, its sensitivity and cardiotoxicity limit DOX’s effectiveness in clinical settings. Thus, there is an urgent need to design regulated and efficient DOX medicine delivery devices. In recent decades, a range of pH-sensitive polymers have been employed for drug administration due to their ability to improve drug solubility, prevent inactivation, and offer controlled release21.

Most tumor tissues are known to have more acidic pH values (pH = 7.4) than healthy tissues. After being moved from healthy cells to cancerous tissues, pH-responsive polymers can change structurally and control the release of therapeutic chemicals. Chitosan and its derivatives are promising candidates for drug delivery as members of pH-responsive polymers because of their robust mucoadhesiveness, biodegradability, biocompatibility, and pH-sensitive properties22. Several studies have evaluated how chitosan’s molecular weight (Mw), degree of acetylation (DD), and sequencing affect how well it delivers drugs23. Studies have shown that doxorubicin’s Mw affects how stable the interaction between it and chitosan is. Researchers found that chitosan’s high Mw and DD can indirectly affect its adhesive properties. However, the molecular processes behind the pH-responsive production of chitosan-DOX complex structures are poorly understood.

The current study was designed to better understand the molecular interactions and the process underlying the absorption and release of DOX drugs from the chitosan–Eudragit carrier. For this purpose, molecular simulations at different sizes have been used to analyze the molecular interactions of this drug delivery system. This work illustrates how AA-MD and CG-MD cause pH-dependent drug release from carrier nanoparticles. This method was used to investigate using chitosan–Eudragit nanocarriers to deliver the anticancer medication DOX. It is one of the earliest in silico studies to show the pH-dependent drug release of chitosan–Eudragit nanoparticles.

Methods and materials

Molecule designs

The simulations of molecular dynamics were carried out as we previously reported24. It can be summed up as follows: Gaussian is used to build the molecular structures of polymers and medications, and Avogadro and Hyperchem tools optimize the structures’ shape. The leading optimization was carried out using Gaussian based on the three-layer ONIOM technique to achieve an accurate electronic structure, particularly for the functional groups critical to binding (e.g., amine and carboxylate groups). The high layer, treating these key groups, used B3LYP/6–311 + G* to ensure an accurate description of anionic oxygen atoms and hydrogen-bonding interactions. The medium and low layers modeled the remainder of the structures to balance accuracy and computational efficiency. The topological parameters use the charge density as calculated by Gaussian. The molecular topologies were obtained using the PolyParGen web service. Next, each molecule in the 6 × 6 × 20 nm3 simulation cells undergoes the leading optimization for 50 ns.

Simulation

To ensure the physical plausibility of the initial configurations and the absence of steric clashes, all systems underwent a rigorous two-stage preparation protocol prior to production simulations. First, an energy minimization was performed using the steepest descent algorithm, which was considered complete only when the maximum force acting on any atom fell below a threshold of 1000 kJ/mol/nm. This critical step relieves any unfavorable atomic overlaps and yields an energetically stable starting structure. Following minimization, the systems were progressively equilibrated. An initial NVT (canonical ensemble) equilibration was conducted to stabilize the temperature, followed by an NPT (isothermal-isobaric ensemble) equilibration to stabilize the system density and pressure. The convergence of these thermodynamic parameters, including potential energy, temperature, and pressure, was monitored to confirm that each system had reached a stable, equilibrated state before commencing the production phase of the molecular dynamics run.

The preliminary simulation was conducted using GROMACS 2020.1 at MD, NPT (constant number of atoms, N; constant pressure, P; constant temperature, T), NVT (constant number of atoms, N; constant volume, V; constant temperature, T), and EM steps. 10 × 10 × 10 nm3 are considered to be simulation boxes with an OPLS-AA force field. The NPT and NVT were performed for 0.5 ns (with 1 fs time step) using the velocity-scaling algorithm at 300 K and the isotropic Berendsen method at 1 bar, respectively, using the MD (100 with 2 fs time step) simulations. The cutoff radius for the Coulomb and van der Waals interactions was 1.2 nm. The temperature and pressure algorithms are the isotropic Parrinello-Rahman method at 1 bar, and the Nose-Hoover (velocity-scaling algorithm in NVT and NPT) at 300 K. We used the Coulomb energy technique and Particle Mesh Ewald (PME). Coarse-grained simulations were performed at different pH levels using the Martini force field in simulation cells 40 × 40 × 40 nm3 for 100 ns (with a 30 fs time step). Table 1 shows further details of the simulation.

Table 1 Summary of simulation system details.
Full size table

The CG-MD system was designed with a larger number of polymer chains and drug molecules within a significantly larger box to better represent the collective behavior and interactions of the nanocarrier system, while the AA-MD system provided high-resolution detail on a smaller, representative complex. The concentration of DOX was calculated based on the number of molecules and the volume of the simulation box.

System protonation at different pH levels

The protonation states of ionizable groups were set according to the simulation pH. For simulations at pH = 5 (acidic), the amine groups of chitosan and doxorubicin were considered fully protonated (-NH3+), and the carboxylic acid groups of Eudragit were considered largely protonated (-COOH). For simulations at pH = 7 (neutral), the chitosan and DOX amine groups remained protonated, while the Eudragit carboxylic acid groups were considered deprotonated (-COO-). These states were explicitly defined in the topology files for the AA-MD simulations and are inherent to the Martini bead types selected for the CG-MD simulations.

Coarse-grained (CG) mapping

The Martini 2.2 force field was employed for coarse-grained molecular dynamics simulations. In this framework, groups of atoms are represented by a single interaction site, or bead, allowing for the simulation of larger systems over longer timescales. The mapping schemes for the key molecules are described below:

  • Chitosan: The glucosamine ring was mapped onto a single polar bead of type SNda, which represents a non-H-bonding donor/acceptor. The protonated amine group (-NH3+) in the glucosamine monomer was represented by a positively charged bead of type Qd.

  • Eudragit: The methacrylate backbone was mapped onto apolar C1 beads. The pendant functional group, critical for pH-responsiveness, was mapped as follows: the carboxylic acid group (-COOH) at pH 5 was represented by a polar P5 bead, while the deprotonated carboxylate group (-COO−) at pH 7 was represented by a negatively charged Qa bead.

  • Doxorubicin (DOX): The anthracycline aglycone core was mapped using semi-polar apolar SC1 beads to represent the aromatic rings. The amine group on the daunosamine sugar was represented by a combination of a charged Qd bead (for the protonated amine) and a polar P1 bead. Hydroxyl and ketone groups in the molecule were mapped to polar P1 and P2 beads, respectively. One DOX molecule was represented by a total of 13 CG beads.

The mapping for all molecules adhered to the standard Martini rule of representing approximately four heavy atoms plus associated hydrogens with a single bead. The interaction parameters between these bead types are defined within the Martini force field and inherently capture the pH-dependent behavior through the assigned charges on the Qd and Qa beads.

Analysis methods

The analysis of all simulation trajectories was conducted using tools within GROMACS 2020.1, supplemented by in-house scripts. To ensure the analysis reflected equilibrated system behavior, an initial equilibration period of 20 ns for the all-atom trajectories and 40 ns for the coarse-grained trajectories was discarded. All subsequent analyses, including the calculation of mean values and their accompanying standard deviations, were performed on the stable production phase of the simulations to ensure statistical reliability.

Structural stability and flexibility were assessed through root-mean-square deviation (RMSD) and root-mean-square fluctuation (RMSF) calculations on backbone and atomic positions, respectively, using the gmx rms and gmx rmsf tools. The compactness of the doxorubicin cluster and the polymer nanocarrier was quantified by the radius of gyration, computed with the gmx gyrate tool, applying the standard Martini bead masses for coarse-grained analyses.

Interfacial interactions were characterized by several methods. Hydrogen bonding in all-atom simulations was identified with the gmx hbond tool using standard geometric criteria of a 0.35 nm donor-acceptor distance cutoff and a 30-degree angle cutoff. For coarse-grained simulations, where explicit hydrogens are absent, persistent polar interactions were analyzed as a proxy by measuring the proximity between polar and charged beads within a 0.55 nm cutoff. The solvent-accessible surface area of the doxorubicin molecules was calculated with the gmx sasa tool to measure hydration, while the non-bonded interaction energies between the drug and carrier were determined using gmx energy. The spatial distribution and density of doxorubicin around the nanocarrier were evaluated with the radial distribution function (gmx rdf), and the interfacial contact area was quantified using a double-distance cutoff method.

System protonation and ion concentration

The protonation states of ionizable groups were manually assigned in the initial topology files for both all-atom and coarse-grained simulations to reflect the specific pH conditions, as a constant-pH molecular dynamics protocol was not employed. The protonation states were determined based on the typical pKa values of the functional groups. For simulations at pH 5, representative of the acidic tumor microenvironment, the amine groups of chitosan and doxorubicin were defined in their protonated state (-NH3+), while the carboxylic acid groups of Eudragit were considered largely protonated (-COOH). For simulations at physiological pH 7.4, the amine groups of chitosan and doxorubicin remained protonated, and the carboxylic acid groups of Eudragit were defined in their deprotonated, anionic state (-COO−). In the Martini coarse-grained model, these protonation states are inherently represented by the selection of appropriate bead types; for instance, charged Qd beads represent protonated amines and Qa beads represent deprotonated carboxylates.

Each system was first neutralized by adding sodium (Na+) or chloride (Cl−) counterions using the gmx genion tool to balance the net charge of the simulation box. Subsequently, a physiological ionic strength of 150 mM was established by adding further NaCl ions to mimic a biologically relevant salt concentration.

Ethics approval

The research was approved by the ethical committee of Tabriz University of Medical Sciences [ethics code: IR.TBZMED.VCR.REC.1404.158].

Consent to participate

This article contains no studies with human participants or animals performed by authors.

Results and discussion

The nanoencapsulation of active pharmaceutical ingredients enhances successful drugs’ pharmacological activity, specificity, tolerability, and therapeutic index25. As was already indicated, pH-sensitive polymers have garnered much interest in TDD because of their remarkable capacity to react to pH changes in various physiological settings. By accurately delivering therapeutic agents to target areas and preventing drug release to off-target regions, this characteristic enables pH-sensitive polymers to function as smart carriers for therapeutic agents, minimizing side effects. Functionalizing nanomaterials and their drug compounds can facilitate drug release at the tumor site at the tumor microenvironment’s slightly acidic pH (approximately 6.5), lower than physiological conditions24. By severing the drug’s bond with the material under the tumor site’s acidic pH, this technique mediates drug release.

Table 2 shows the mean, minimum, and maximum values of the RMSD and root-mean-square fluctuation (RMSF) in the specified collection. Based on this data, it was found that the average RMSD at neutral pH was lower than that at acidic pH, and the range of oscillations between the maximum and minimum RMSD at neutral pH was narrower than that at acidic pH. This study shows that the average volatility of all atoms is lower in the neutral pH condition than in the acidic state. These findings also demonstrate that the RMSF changes are less at both the maximum and minimum at neutral pH than at acidic pH. According to these assessments, the system and simulation are getting closer to a steady state; the system is more stable in the neutral state, and this stability level is higher than in the acidic condition.

Table 2 The RMSD, RMSF, and DOX mean, maximum, and minimum values.
Full size table

Results are displayed more accurately but on smaller scales in an AA-MD simulation. The charge density and simulation box (with the volume unit in coulombs/box) are shown in the first study; the charge distribution analysis was constructed for DOX molecules, chitosan, and eudragit monomers (Fig. 1). Figure 1a shows the charge distribution for the eudragit monomer. Because the polymer’s charge is negative and the ratio of negative charges is greater than that of positive charges, this monomer has a proportionate number of positive-negative charges. Furthermore, as seen in Fig. 1b, the chitosan monomer has a distinct charge distribution. This monomer contains positive and negative charges; the amine group controls the positive charge, and the OH group controls the negative charge.

The DOX molecule has a comparatively positive charge, as shown in Fig. 1c. Because of its non-portable amino group, DOX has a positive charge at pH 7.0. As a result, it can engage in electrostatic interactions with drug carrier groups that are totally or partially negatively charged. Experimental research has demonstrated that chitosan-based nanoparticles have a high encapsulation effectiveness and that the protonation of chitosan’s amino groups can result in pH-sensitive DOX release.

Fig. 1
figure 1

The charge distribution in the structures of the molecules (a) eudragit, (b) chitosan, and (c) DOX was ascertained using the AA-MD simulation technique.

Full size image

Figure 2 also shows the energy and number of hydrogen bond diagrams for the chitosan and eudragit polymers at different pH values and the DOX molecule. The energy diagram is calculated using the electrostatic van der Waals energy and the total energy (Fig. 2a and b). The Leonard Jones equation calculates van der Waals energy from molecular interactions. This equation uses various atomic weights and molecules. The calculation of electrostatic energy also depends on each atom’s charge, which is determined by Coulomb’s equation.

Every atom in the AA-MD simulation has a mass and a charge. The total energy is the sum of the van der Waals and electrostatic energies. Energy analysis is essential in pH-dependent simulations because the charge of the atoms and the number of hydrogens and proteins in the system change with pH26. The immediate changes in electrostatic and total energy caused by the shift in charge affect van der Waals’s energy. Figure 1’s graphs demonstrate how the number of protons rises with system acidity. Its electrostatic energy increases as a result. At pH = 7 (neutral), the average total energy is about − 1200, while at pH = 5, the charge is positive, the number of protons increases, and the average total energy is about 7000. However, there has been a total change in the van der Waals energy, which is dependent on the mass of the atoms. These results indicate that DOX desorption occurs at an acidic pH, while DOX adsorption through chitosan and eudragit nanocarriers occurs at a neutral pH. Therefore, it may be claimed that DOX excretion also affects the cancer tumor’s acidic pH.

Another important consideration in pH-dependent simulations is the examination of hydrogen bonds, which have greater strength than van der Waals and electrostatic energies. Figure 2c illustrates the changes in pH as the number of hydrogen bonds increases across the simulation time. Protonation also requires changes in pH. More hydrogen bonds are formed at a neutral pH than an acidic pH. According to Arrhenius’ theory, when cancer cells are in an acidic environment, the drug and nanocarrier molecules release themselves into the hydrogen environment because of the decreased average hydrogen bonding at acidic pH27. Consequently, the molecules acquire and lose the same charge at an acidic pH.

Fig. 2
figure 2

The interactions between DOX and the nanocarrier are displayed on the energy graph of the AA-MD simulation approach at (a) pH = 5 and (b) pH = 7. It also shows the hydrogen bonds between DOX and the nanocarrier at (c) pH = 5 and (d) pH = 7.

Full size image

Figure 3 shows the radial distribution function (RDF) and radius of gyration graphs. The distribution of the RDF graphic displays the DOX charges surrounding the nanocarrier at different pH levels. The graph’s maximum shows the greatest absorption rate of the nanocarrier and DOX molecules. DOX molecules were more adsorbed at neutral pH, as seen in Fig. 3a. It is also evident that fewer DOX molecules are at neutral pH at distances larger than 2 nm. This plot demonstrates that at neutral pH, as opposed to acidic pH, DOX molecules are more densely packed around the nanocarrier molecules. Figure 3b also shows the radius of gyration, which is the radius of the stacked molecules at different simulation times. This graph shows that DOX molecules and nanocarriers are more aggregated at neutral pH. This graph shows that the average radius of gyration and its endpoint at neutral pH are smaller than those at acidic pH. Simulation results show that at pH neutrality, DOX molecules are more adsorbed.

Fig. 3
figure 3

(a) The concentration of DOX drug surrounding the nanocarrier at pH values of 5 and 7; (b) The gyration radius of DOX drug changes over time at pH values of 5 and 7 using the AA-MD simulation method.

Full size image

Figure 4 shows the data obtained from the solvent-accessible surface area (SASA) research. The water surface area found in DOX molecules is shown in Fig. 4a. At an acidic pH, DOX molecules came into more contact with the water. The graph’s steady decline at neutral pH also suggests that DOX molecules are continuously adsorbed on the nanocarrier at that pH. Simultaneously, the graph exhibits several oscillations under acidic circumstances. These changes could be brought on by the formation of electrostatic repulsion in the acidic state. Figure 4b also shows how the contact area between the nanocarriers and DOX molecules changes over time. According to the contact surface area, the drug-nanocarrier contact surface in the neutral state is approximately three times larger than the acidic state’s. As a result, the neutral state had substantially higher DOX molecule adsorption.

Fig. 4
figure 4

(a) Interaction of molecules during the simulation. (b) The contact surface of the water and DOX changes when they interact with the nanocarrier, and (c) the contact surface of nanocarrier particles at pH = 5 and pH = 7 changes using the AA-MD simulation approach.

Full size image

Figure 5 also shows the RMSD and RMSF plots. These graphs display the system’s oscillations over time, one for each atom. The oscillations are higher at the start of the simulation and lower at the end. This pattern suggests that stability is approaching for the system. In other words, as the simulation duration increases, the molecules’ energy level falls. Table 2 also provides the average, maximum, and minimum RMSD and RMSF values. Figure 5a; Table 2 show that the average RMSF at neutral pH is lower than at acidic pH. RMSD analysis, which displays the identical data for each atom in Fig. 5b, indicates that the range of oscillations between the maximum and minimum RMSD at neutral pH is lower than at acidic pH.

This investigation revealed that the neutral pH state has lower average volatility for all atoms at all time intervals than the acidic state. Additionally, Table 2 shows that the neutral pH has smaller maximum and lowest RMSF swings than the acidic pH. These evaluations show that the system and simulation are getting closer to a steady state, that the neutral condition makes the system more stable, and that this stability level is higher than that of the acidic state.

Fig. 5
figure 5

Using the AA-MD simulation method, (a) DOX atom oscillations during interaction with a nanocarrier at pH values of 5 and 7 and (b) DOX particle oscillations over time at pH values of 5 and 7.

Full size image

In addition to the AA-MD simulation approach, CG-MD simulations were performed. These simulations were conducted at greater scales and used more molecules than the AA-MD simulation. Figure 6 displays the energy graphs derived from CG-MD analysis. The outcomes of these plots are nearly identical to those from the AA-MD approach. These findings indicate that adsorption occurs at neutral and acidic pH values and that molecules repel one another due to electrostatic energy. Figure 7a also displays the radius of gyration and SASA plots produced by CG-MD analysis. Although the radius of gyration in this graph is in a higher range than that of the AA-MD technique, the results fully validate that study. Based on the findings of the radius of gyration investigation, the aggregation occurred in the neutral state.

Fig. 6
figure 6

The interactions between DOX and the nanocarrier over time at (a) pH = 5 and (b) pH = 7 are depicted in the energy graph generated by the CG-MD simulation method.

Full size image

Figure 7b displays the results of the CG-MD analysis we performed on the SASA plot. This analysis ascertained DOX’s neutral state adsorption. These results demonstrated that DOX molecules were attached to the nanocarrier and did not float in water because they had a smaller contact surface with water at neutral pH. The simulated snapshots are displayed in Fig. 7, where the molecules interact in the microstate and are widely dispersed because the scale of the molecules in the CG-MD mode is somewhat outside of the nano-state.

Figure 7c also displays the contact region between DOX and nanocarriers. This figure illustrates how the contact area between medications and nanocarriers increases at neutral pH. However, this contact area doesn’t change much over the simulation at an acidic pH. Additionally, during the simulation, the repulsions are near to one another. According to the experimental research done thus far, chitosan is a promising agent for creating pH-responsive nanocarriers for the transport of DOX and the suppression of cancer. Other ligands also easily trigger it to increase the selectivity of nanoparticles against cancer cells. Figure 7d shows the simulation procedure at the start and ending points.

Fig. 7
figure 7

The DOX gyration radius changes over time at pH = 5 and pH = 7 through CG-MD simulation; the DOX and water contact surfaces change when they interact with the nanocarrier at pH = 5 and pH = 7; and the contact surface changes between nanocarrier particles at pH = 5 and pH = 7 are the results of CG-MD simulation.

Full size image

According to recent research, mitogen-activated carbon nanocomposites based on chitosan/polyvinylpyrrolidone (PVP) release DOX in a pH-sensitive manner, simulating the pH of the tumor microenvironment and triggering apoptosis to lower breast cancer survival dramatically. According to a recent experiment, graphene-chitosan nanocomposites for the delivery of DOX have been created and stabilized using BSA. The presence of BSA prevents the drug’s explosive release from the chitosan nanocomposites. They can release the medication for up to 28 days (84%) and exhibit consistent release over 24 h. According to in vivo research, the nanocarrier’s tumor penetration ratio was 83.5%, whereas free DOX’s was 46.1%.

Eudragit, a synthetic polymethacrylate copolymer used in this work, has a wide range of hydrophilicity and hydrophobicity, a variable tolerance to functional groups, and the ability to cross-link under different conditions12. Therefore, the gold standard for employing drugs to treat various ailments is eudragit polymer grades. Studies have shown that by combining different grades, the long-term release of eudragit-harmless synthetic polymers-can be tailored to match certain requirements28.

Chitosan’s major amine functional group is responsible for its exceptional biological qualities, which include mucoadhesiveness, regulated drug delivery, and improved penetration29. Chitosan’s positively charged ability to connect to various mammalian cells makes it useful for tumor tissue selection. Furthermore, because chitosan dissolves in acidic solutions, protonated chitosan, and its derivatives can be added to anticancer medications poorly soluble in water to improve their solubility and biocompatibility. The created chitosan–Eudragit nanocarriers are expected to be suitable for properly targeted drug delivery, and more research is advised.

Conclusions

This study employed integrated all-atom and coarse-grained molecular dynamics simulations to elucidate the pH-dependent mechanism of doxorubicin (DOX) release from chitosan–Eudragit nanocarriers. The key mechanistic insight is that the protonation state of the nanocarrier’s functional groups directly governs its interaction with the drug. At physiological pH (7.4), the deprotonated carboxylate groups of Eudragit and the protonated amine groups of chitosan facilitate strong electrostatic attraction and hydrogen bonding with DOX, leading to a stable complex characterized by a large contact area and minimal fluctuations. In contrast, the acidic tumor microenvironment (pH 5.0) triggers the protonation of Eudragit’s carboxylate groups, inducing significant electrostatic repulsion against the similarly protonated amines of chitosan and DOX. This repulsion destabilizes the nanocarrier, drastically reduces drug-carrier interactions, and ultimately drives the controlled release of the therapeutic payload.

The primary practical implication of these findings is that chitosan–Eudragit nanocarriers are a highly promising platform for the targeted delivery of anticancer drugs like DOX. Their inherent biocompatibility, combined with this demonstrated pH-responsive smart release mechanism, positions them as effective candidates for minimizing off-target toxicity and enhancing therapeutic efficacy.

Future work should focus on experimentally validating these computational predictions through in vitro drug release studies and cytotoxicity assays. Furthermore, employing more advanced simulation techniques, such as constant-pH molecular dynamics, could provide a more dynamic description of the protonation events. Finally, exploring the microfluidic synthesis of these nanocarriers and their evaluation in in vivo models would be essential steps toward translational application. Therefore, this computational study not only clarifies a fundamental mechanism but also provides a robust foundation for the rational design and development of next-generation, stimuli-responsive drug delivery systems.