Theoretical and Experimental Insights into the Possible Interfacial Interactions between β-Glucan and Fat Molecules in Aqueous Media

Abstract

Excessive body fat and high cholesterol are one of the leading reasons for triggering cardiovascular risk factors, obesity, and type 2 diabetes. Beta-glucan (BG)-based dietary fibers are found to be effective for lowering fat digestion in the gastrointestinal tract. However, the fat capturing mechanism of BG in aqueous medium is still elusive. In this report, we studied the dietary effect of barley-extracted BG on docosahexaenoic acid (DHA, a model fat molecule) uptake and the impact of the aqueous medium on their interactions using computational modeling and experimental parameters. The possible microscale and macroscale molecular interactions between BG and DHA in an aqueous medium were analyzed through density functional theory (DFT), Monte-Carlo (MC), and molecular dynamics (MD) simulations. DFT analysis revealed that the BG polymer extends hydrogen bonding and nonbonding interactions with DHA. Bulk simulation with multiple DHA molecules on a long-chain BG showed that a viscous colloidal system is formed upon increasing DHA loading. Experimental size and zeta potential measurements also confirmed the electrostatic interaction between BG–DHA systems. Furthermore, simulated and experimental diffusion and viscosity measurements showed excellent agreement. These simulated and experimental results revealed the mechanistic pathway of how BG fibers form colloidal systems with fat molecules, which is probably responsible for BG-induced delayed fat digestion and further halting of fatty molecule absorption in the GI tract.

Graphical Abstract

INTRODUCTION

Recently, the world has seen a phenomenal rise in cardiovascular, cancer, diabetes, and other deadly diseases.^1,2 Cardiovascular and diabetes diseases are two of the most common and deadly diseases in the United States associated with obesity.^2,3 Because of this, regardless of the significant side-effects associated with the anti-obesity medications, several drugs such as orlistat (Xenical), phentermine and topiramate (Qsymia), bupropion and naltrexone (Contrave), and liraglutide (Saxenda) received food and drug administration (FDA) approval.^4,5 Biocompatible materials that interact with fat molecules and prevent them from being transported through the GI tract have the potential to be used as therapeutic modalities to treat diet-induced obesity. For instance, we have recently developed an ionic liquid, composed of choline and geranic acid, that interacts and traps fat molecules within the intestine when given orally. We have observed that 10 μL of daily oral dose of the ionic liquid in a high-fat-diet rat model resulted in the reduction of body weight by 13% over 4 weeks of treatment.⁶

Beta-glucan (BG) is a naturally derived polysaccharide composed of soluble dietary fibers with three β (1–4) and one β-(1–3) glycosidic-linked glucose molecules in a repeating unit (RU).⁷ Due to its diverse physiological functions, bioactive (biocompatibility and biodegradability) properties, and health benefits, BG has been widely exploited in pharmacology and targeted drug delivery.^8,9 BG has shown to have an affirmative action in controlling glucose levels and serum cholesterol and attenuating coronary heart disease conditions.^10–13 A significant number of clinical studies and consequent meta-analysis have been carried out that strongly support oat/barley BG’s efficiencies in lowering LDL (low-density lipoprotein) cholesterol in the human body.^11,14 Consequently, leading regulatory agencies like the United States FDA have issued the health benefits of oat/barley BG and its functionality in mitigating blood lipid concentrations and improving conditions responsible for metabolic syndrome.^15,16 It has been reported that the BG polymer utilizes its viscosity property and gel-forming ability to form droplets with fat molecules.¹⁶ Along the same line, BG fibers help increase the viscosity and structural rigidity of the lipid droplets, which further delays the lipid digestion process by inhibiting the lipase enzyme activity in the small intestine.^11,16 A recent study by Thandapilly et al. demonstrated that BG-mediated hypocholesterolemia largely depends on the molecular weight (MW) of the BG and occurs through an increase in the bile acid excretion and short-chain fatty acids in the intestine.¹⁷ However, elucidating the working mechanism of how BG interacts with the fat molecules is still difficult because of the structural complexity and variability among the resources.

Computational modeling is an emerging technology that helps with the analysis of molecular structures, the rationalization of structural properties, and the assessment of interaction efficiencies beforehand in an environment that can mimic the experimental conditions. The results of the computational analysis can be employed in optimizing the experimental design. In recent studies, computational simulations based on density functional theory (DFT), semi-empirical theory, and molecular dynamics (MD) simulation have demonstrated the molecular interaction of many polymeric nanoparticles and anticancer drugs.^18–20 DFT simulations can be utilized to investigate the microscale molecular interaction processes through analyzing the COSMO profile, frontier molecular orbital interactions, electron density, and charge distribution profiles, while MD simulations can shed light on the molecular interaction in the bulk system.^19,21,22 MD can be utilized to simulate the interactions of BG molecules with other biologically significant molecules in the aqueous system. Furthermore, MD simulation can be used to study bulk dynamic and structural properties like intermolecular distances radial distribution function (RDF), bond length distribution, mean square displacement (MSD), and diffusion coefficients.^18,19,23,24 Combining the results from DFT and MD simulations could provide a better understanding of how BG molecules interact with various biomolecules in the presence of water.

In this work, we have investigated the dietary effect of barley BG on fat uptake. Barley BG fibers are relatively less viscous, water-soluble, and are often beneficial due to their multiple bioactive functionalities, including cholesterol resistivity and anti-obesity. Docosahexaenoic acid (DHA), a well-studied omega-3 fatty acid, is considered a model fat molecule due to the low complexity of its structure. We simulated the structures of BG and DHA in aqueous systems to demonstrate and understand the physical and chemical interaction between these molecules. We observed that water molecules influence interaction between BG and DHA and assist in the formation of a colloidal system. Monte-Carlo (MC) and MD simulations revealed the effects of increasing DHA loading (BG/DHA- 1:5, 1:7, 1:10, 1:12, 1:15, and 1:20) on its adsorption on the BG polymer and on corresponding changes in molecular interactions based on the variation in their dynamic and structural properties in an aqueous system. The interactions between BG and DHA were stable over 10 nanoseconds (ns) of MD simulation and were highly influenced in the presence of water. Experimental setups with similar BG/DHA ratios were prepared for size, zeta potential, diffusion, and viscosity measurements. In both simulated and experimental setups, the BG–DHA-absorbed systems demonstrated increased viscosity with DHA loading, indicating that BG can form a colloidal system upon interaction with DHA. The findings from this study can help extend the understanding of how dietary BG fibers undergo structural rearrangements to form a colloidal solution with fat molecules in the GI tract.

MATERIALS AND METHODS

DFT, MC, and MD Simulation Methodology.

We used DFT- and force field-based semi-empirical theories for the analysis. BG, DHA, and water molecules were considered for different computational simulations, optimizations, and molecular interaction analyses. These simulations were carried out to obtain information regarding molecular interactions (hydrogen bonding, van der Waals, and electrostatic), spectral analysis, highest-occupied molecular orbital (HOMO)–lowest-unoccupied molecular orbital (LUMO) distributions, MD, and bulk particulate formation. For molecular interactions and orbital and spectral analysis, we considered the structure of a single RU of BG originating from barley flour and its interaction with a single DHA fat molecule by placing them in the COSMO-based aqueous solvent system. Geometry optimization, energy, frequency, and molecular interactions were analyzed with the DFT-based DMol3 module (Materials Studio DMol3 version 2017). Although for bulk analysis, the interaction of a five RU-long BG molecule obtained from a similar barley source with different DHA loadings (5–20 molecules) and 1000 molecules of water was considered. We used condensed-phase-optimized molecular potentials for atomistic simulation studies-II (COMPASS-II) force field for this study.

DFT analysis was carried out using Perdew–Burke–Ernzerhof generalized gradient approximation functionals. Spin was set as unrestricted, with the formal spin as initial. For convergence, the custom parameters were set as energy, 1.0 × 10⁻⁵ Ha; maximum force, 0.005 Ha/Å; maximum displacement, 0.01 Å; and maximum step size, 0.3 Å. The double numerical plus polarization basis set was used with DFT semi-core pseudopots for core treatment for analyzing the electronic options. The COSMO-based water solvent model was used for all DFT-based simulations. The self-consistent field tolerance was 0.5 × 10⁻⁴. The predefined DFT-D parameters were used throughout the calculations.

For simulating the bulk system, a cubic vacuum slab was generated using 10 × 10 × 10 nm³ vacuum space. The five-RU BG molecule was placed in the middle of the vacuum chamber. For the MC and MD simulations, all systems were prepared with 1000 H₂O with various BG/DHA (1:5, 1:7, 1:10, 1:12, 1:15, and 1:20) loadings. The MC simulation process was utilized in determining the adsorption sites, while MD simulation was used for finding out the dynamic and structural properties of the BG–DHA–H₂O systems. The COMPASS-II force field was used in these studies. The parameters for geometry optimization with the COMPASS-II force field for energy, force, and displacement were 2.0 × 10⁻⁵ kJ/mol, 0.001 kJ/mol/Å, and 2.0 × 10⁻⁵ Å, respectively. The aqueous-system MD simulations were carried out initially in the isothermal–isochoric (NVT) and then in the isothermal–isobaric ensemble (NPT) for 10 ns time so that the BG–DHA–H₂O systems could reach equilibrium.

MATERIALS AND CHEMICALS

Less viscous barley BG (~179 kD) was purchased from Megazyme, USA (Cas. no.: 9041-22-9). DHA was purchased from Nu-Check Prep, Inc. USA. (Cat. no.: U-84A-MA6-E). Simulated gastric juice with 0.2% NaCl in 0.7% (v/v) HCl (Ricca chemical) and rhodamine B (RhB) fluorescent dye were purchased from Thermo Fisher Scientific. 1-Ethyl-3-(3′-dimethylaminopropyl)carbodiimide hydrochloride (EDC HCL) and 4-dimethylaminopyridine (DMAP) were also purchased from Thermo Fisher Scientific. All other chemicals were purchased from Thermo Fisher Scientific, USA, unless mentioned otherwise. Milli Q water was used throughout the experiments.

Sample Preparation.

Based on the systems created for MD simulation (Supporting Information Table S2), six BG–DHA solutions were experimentally prepared. First, 10 mg of BG was dissolved in 500 μL of water through continuous heating and stirring for 20 min. The resulting solution was observed to be transparent. Five more BG solutions were prepared following the same procedure and placed on a hot plate. Next, 5, 7, 10, 12, 15, and 20 mole ratios of DHA with respect to the moles of BG (10 mg) were calculated and added dropwise to six glass vials containing BG solution under stirring condition. The solutions were continuously stirred at room temperature overnight to ensure complete mixing of DHA, BG, and water. The solutions turned cloudy after this step. Next, the cloudy solutions were transferred separately into 1 mL plastic tubes each and centrifuged at 1000 rpm for 8 min. Due to centrifugation, all unbound DHA molecules were separated and deposited in the upper layer of the supernatant. Next, the precipitates obtained were redispersed in 500 μL of water. The final concentrated dispersed solutions were used for size, zeta potential, and diffusion coefficient measurement.

Size, Zeta Potential, and Diffusion Coefficient Measurement.

The particle size, distribution, diffusion coefficient, and zeta potential of each sample were measured using a dynamic light scattering (DLS) instrument (Malvern instruments Ltd, Zetasizer-nano ZS series). 4 μL of the sample was dispersed in 1 mL of water and used for all measurements. All measurements were carried out under room temperature conditions. The viscosity of the sample solution was calculated using the value of size and the diffusion coefficient using the Stocks–Einstein equation.

In Vitro Intestinal Permeation Study.

Sample Preparation.

RhB dye-conjugated DHA and BG–DHA were utilized for the in vitro permeation study. Briefly, 0.073 g of RhB and 0.064 g of DMAP were dissolved in 2 mL of anhydrous DCM solution under RT condition. After 20 min of stirring, 0.15 g of EDC.HCl was added to the mixture and gently mixed at 4 °C for 1 h. After that, 100 mg of DHA was added to the reaction medium and stirred vigorously at 4 °C for 48 h to ensure complete conjugation of RhB with DHA. Next, the solution was centrifuged at 3000 rpm for 15 min. The organic supernatant containing unbound RhB was discarded, and the precipitates were washed three times with aqueous solvent to ensure all unbound RhB being removed from the sample. Later, the washed conjugated DHA–RhB sample was dispersed in 2 mL of water and stored at 4 °C until used.

To prepare the BG–DHA–RhB composite, 50 mg of BG was weighed and dissolved in 1 mL of water through vigorous stirring at 60 °C for 20 min. After dissolving BG, 500 μL of dissolved BG was added to DHA–RhB solution and mixed overnight at 4 °C. After mixing, the solution was subjected to centrifugation at 1000 rpm for 10 min, and the precipitate obtained was washed repeatedly to remove any unbound DHA–RhB from the mixture. The final conjugated BG–DHA–RhB was dispersed in 2 mL of water for the diffusion study.

Preparation of the Intestine.

The in vitro intestinal permeation study was carried out using the porcine intestine purchased from a local meat shop and stored at −20 °C until used. Initially, the samples were first thawed at 4 °C for 8 h and later allowed to equilibrate at RT for 20 min and to make them soft enough for cutting into small blocks. While opening the intestinal tract, extra precautions were taken to prevent any damage to the inner mucous layer of the small intestine. Then, the open intestinal tract was cut into appropriate square blocks (1.77 cm²) suitable for mounting on the automated diffusion cells.

Instrumentation.

An ILC07 automated flow-through system (PermeGear Inc., Hellertown, PA, USA) associated with seven in-line diffusion cells was used for this study. The intestinal blocks were tightly screwed in place on each of the seven diffusion cells using adjustable knots and 200 μL of the DHA–RhB/BG–DHA–RhB formulations were loaded on top of the blocks. Note that the inner layer of the intestine was faced upward. Each of the diffusion cells has an inlet channel for letting PBS flow underneath the cell containing the intestinal part and an outlet channel for carrying the PBS + formulations (penetrated through the intestine) out of the blocks. The other side of the inlet channels was connected with a multi-channel peristaltic pump IPC (Ismatec, Zurich, Switzerland), which draws PBS from a reservoir and allows it to flow into the channels at a specific flow rate (4 mL/h), while the liquid discharged out of the outlet channels was collected in 20 mL scintillation vials at seven different time periods (30 min, 1, 2, 3, 4, 5, and 6 h), respectively. All cells were connected with a Julabo BC4 circulating water bath (Seelbach, Germany), which ensured a temperature of 37 °C similar to body temperature throughout the experiment. The samples obtained after intestinal permeation were analyzed by an endpoint UV–vis absorbance study at a wavelength of 550 nm using a Synergy H1 96-well plate reader.

RESULTS

DFT-Based Simulations.

Structure optimization, characterization, and molecular interaction analysis were carried out using DFT-based computational simulation. The structure of DHA and RUs of BG were optimized using the DMol3 module. Their structures were confirmed by comparing the infrared (IR) spectra with the experimentally observed ones reported elsewhere.^7,25 After structural verification and characterization, the molecular interactions between the optimized BG and DHA were investigated. For this, changes in several properties, including IR spectrum, sigma profile, electron density, HOMO–LUMO distribution, and partial density of states (PDOS), were analyzed before and after any molecular interactions between BG and DHA. All molecular analyses relating to DFT calculations and corresponding discussions are presented in the following sections.

Structure Optimization and Characterization by DFT.

Figure 1A shows the DFT-optimized 3D atomistic structure of a RU of barley-extracted BG, demonstrating one β-(1–3)- and three β-(1–4)-glycosidic bonds among α-D glucose molecules. Similarly, the structure of DHA (standard omega-3 fatty acid) was optimized using the same geometry optimization parameters used for BG optimization (Figure 1B). Figure 1D–F represents the corresponding IR spectra of the simulated BG and DHA molecules. The respective IR bands of both BG and DHA are assigned based on the experimental IR reported for both barley BG and DHA omega-3 fatty acid.^7,25 The data for both simulated and experimental IR results for BG and DHA are given in Table S1.

Figure 1.

DFT simulation of molecular structures and corresponding IR spectra. Figure A, B, and C show DFT-based optimized structures of BG, DHA, and BG–DHA system, respectively. Figures D, E, and F show the corresponding simulated IR spectra of BG, DHA, and BG–DHA systems.

The simulated and the experimental BG showed almost similar IR absorption bands typical for polysaccharide structures.⁷ Standard bands observed are: strong absorption at 3000–3500 cm⁻¹ associating O–H stretching vibration, weak C–H (methyl) extending vibration at 3000–2850 cm⁻¹, strong C–H (methylene) and weak C–H (methyl) bending vibrations at 1500–1465 and 1450–1300 cm⁻¹, strong ring vibration converging C–O stretching vibration at 1200–1000 cm⁻¹, and a relatively weak band around 885–895 cm⁻¹ indicating glycosidic linkage (Table S1). Similarly, IR bands of both the simulated and the experimental DHA were compared and showed a good match.²⁵ For this, observed bands are: relatively weak O–H stretching vibration at 3500–3100 cm⁻¹, strong ═C–H (alkene) and C–H (methylene) stretching vibrations at 3100–3006 and 3000–2850 cm⁻¹, respectively, strong C═O (carboxylic) stretching vibration at 1750–1770 cm⁻¹, strong C–H (methylene and methyl) bending vibration at 1465–1300 cm⁻¹, and C–O (ester) stretching and ═C–H bending vibrations at 1200–1000 and 890 cm⁻¹, respectively. These similarities in IR spectral bands indicate that the structural properties of the simulated BG and DHA are very much identical with the experimental results. After assessing the molecular structure of BG and DHA, both structures were simulated together using the identical parameters.

Figure 1C of the optimized BG–DHA system displays hydrogen bonds (blue dashed line, distance between sites 2.5 Å) and close-contact interaction (purple dashed line, distance between sites 3.0 Å). The close-contact interaction refers to van der Waals and electrostatic-type interactions. Notably, BG itself shows intramolecular hydrogen bonding among the hydroxyl functionalities. The IR spectrum displays comparatively strong C–H (methylene) stretching vibration at 2950–2825 cm⁻¹, the characteristic absorption band at 1670–1600 cm⁻¹ for weak C═C stretching vibration, relatively strong C–H (methylene) bending vibration at 1470 cm⁻¹, and other similar bands discussed above. Interestingly, a characteristic absorption band associated with O═C═O stretching vibration was observed at 2396 cm⁻¹ and was absent in their individual spectrum. Also, the O–H absorption band was not observed in the conjugated BG–DHA system. This indicates that the O–H groups were somehow occupied or bound in the conjugated system for which the stretching vibrational band for the O–H group was found to be absent in the spectrum. One possibility for obtaining the O═C═O band and subsequent disappearance of the O–H band could be due to the dominating close proximal interaction between the O–H group of DHA and the H–O–C group of BG, which probably was responsible for this absorption band. Additionally, the C–H group’s enhanced stretching and bending vibrations further confirm that BG and DHA interact in the conjugal system. BG and DHA are likely to interact efficiently through hydrogen bonding and van der Waals/electrostatic interactions in the aqueous solvent system. Therefore, to understand the effect of solvent on BG–DHA interaction, we further analyze the COSMO profile in the next section.

Molecular Interaction Studies through the Sigma Profile.

COSMO utilizes solvation thermodynamics for studying the molecular interactions through analyzing the electronegativity and electropositivity of the target species.²¹ In this system, the molecules of interest are placed within a cavity of the solvent. The molecules in this cavity draw opposite charges from the solvent based on its charge and dipole moments.²⁶ The computational simulation determines the surface charge distribution of the target molecule based on the amount and type of charge the target molecule takes from the solvent system. Hence, based on the COSMO profile similarities of the different molecular systems, it is possible to predict the effects of the solvent charge distribution on the BG-DHA conjugal interactions. Figure 2A shows the COSMO profile for H₂O. There are two peaks around −0.012 and −0.007 e/Ų, which might be due to the lone pairs of oxygen (O) atoms in the H₂O molecule.²¹ The peak around 0.014 e/Ų is likely due to the positive dipoles of the two hydrogen (H) atoms. The peaks around 0.001 e/Ų are indicative of the nonpolar regions of the H₂O molecules.

Figure 2.

DMol3-based COSMO sigma profile. The sigma profile plots of H₂O, BG, DHA, and BG–DHA are shown in Figures A–D. The inset figures show the COSMO surface profiles.

The electron density distribution profile of H₂O (Figure S1) matches well with the predictions from the COSMO profile. Figure 2B shows the COSMO profile for BG. The observed peaks are around −0.003, −0.006, 0.001, 0.005, 0.008, and 0.01 e/Ų, respectively. The peaks around −0.006 and −0.003 e/Ų are likely due to the H atoms polarized by the negative charge-bearing O atoms of the C–OH groups of the α-D glucose molecules.²² The peak around 0.001 e/Ų is for the nonpolar region, while the peaks around 0.005 and 0.008 e/Ų might be due to the interaction between the C and O atoms. The nonpolar H atoms of the CH₂ are responsible for the charge density around 0.01 e/Ų. The peak around −0.003 e/Ų of DHA is also likely due to the polarized H atoms. The peak around 0.005 e/Ų is indicative of the interaction between C and O atoms, and the nonpolar H atoms are indicated by the 0.01 e/Ų peak.

The BG–DHA system shows peaks around −0.004, −0.003, 0.001, 0.004, and 0.012 e/Ų, respectively. The peak values indicate C–OH, C–O, nonpolar H atoms, and nonpolar regions for the BG–DHA system.^22,26 A slight shift in the peaks for BG–DHA indicates the change in the electronic charge density due to interaction between BG and DHA molecules. The sigma profile of BG–DHA matches well with that of BG and DHA, with little resemblance to the sigma profile of water. The electron density profiles from Figure S1 also show that there is slight overlap between the electron clouds of BG and DHA. This demonstrates that the BG and DHA molecules are more likely to interact with each other in the presence of an aqueous solvent.

Molecular Orbital and Electronic Structure Analysis.

The IR and COSMO plot analysis showed the possibility of interaction between BG and DHA in an aqueous system. In this section, the frontier molecular orbitals, that is, HOMO–LUMO, and projected density of states (PDOS) are analyzed to see how the orbital and electronic structures varied when BG and DHA were taken together. Figure 3A shows that both the HOMO (−0.207 eV) and LUMO (−0.189 eV) are distributed around the β-(1–4)-glycosidic bond furthest from the β-(1–3)-glycosidic bond for the BG RU. In the case of DHA, the HOMO (−0.188 eV) is distributed throughout the whole hydrophobic chain, while the LUMO (−0.035 eV) is distributed around the –COOH group (Figure 3B). Figure 3C shows the HOMO–LUMO distribution for the optimized BG–DHA conjugal structure, demonstrating that there is significant redistribution of the HOMO (−0.19 eV) and LUMO (−0.184 eV). Note that the HOMO–LUMO distribution and corresponding energies of all systems are obtained in the COSMO-based solvent system. This indicates that there is an interaction between BG and DHA. The sites where the HOMO and LUMO reside are also the sites where BG and DHA form the hydrogen bonds and close contact interaction. It is obvious that the interaction is strictly non-electron transfer, as both the HOMO and LUMO are distributed over the BG RU. To further analyze what type of interactions occur between BG and DHA, the binding energy (BE) $\left(E_{\text{b}}\right)$ value was calculated according to eq 1

$$E_{\text{b}}=E_{\text{BG}-\text{DHA}}-\left(E_{\text{BG}}+E_{\text{DHA}}\right)$$ (1)

Figure 3.

Frontier molecular (HOMO–LUMO) orbitals distribution. (A–C) HOMO–LUMO distributions for BG, DHA, and BG–DHA.

Here, the energy values of the optimized BG–DHA, BG, and DHA systems were found to be of $E_{\text{BG–DHA}}(-3523.3151051\text{Ha})$, $E_{\text{BG}}(-2516.3523939\text{Ha})$, and $E_{\text{DHA}}(-1006.9494163\text{Ha})$, approximately. Thus, according to eq 1, the BE, E_b value, for the conjugated BG–DHA system should be −0.0332949 Ha or −34.90 kJ/mol. This energy value shows that BG and DHA interaction is feasible and produces a stable BG–DHA system.

The electronic structures of BG, DHA, and BG–DHA systems were analyzed based on Mulliken charge distribution (Figure S2) and the PDOS (Figure S3).^27,28 By comparing the PDOS of BG and DHA with that of the BG–DHA system, it is possible to analyze how these two molecules might interact at the microscopic level. Figure S3 shows the simulated PDOS for the BG, DHA, and BG–DHA systems. Like the findings from the HOMO–LUMO analysis, there is a minimal band gap in the electronic structures of BG, DHA, and BG–DHA. The p bands originate from the O 2p orbitals, while the s bands are from both the H and O atoms. In BG, the s and p bands overlap significantly between −8 to −16 eV and 4 to 8 eV (Figure S3A). For the DHA molecule, the bands overlap around −8 to −12 eV and partially overlap around 4–8 eV (Figure S3B). Figure S3C shows that the s and p bands overlap around −8 to −12 eV and partially overlap around 4–8 eV for the BG–DHA system. The Mulliken charge distribution shows (Figure S2) that the atoms of the BG–DHA system that took part in the hydrogen bonding and close-contact interaction had different charges compared to the individual BG and DHA systems. However, only partial changes in charge of the atoms indicated that the interaction was preferentially nonbonding.

Bulk Simulation.

For the bulk simulation, COMPASS II force field of MC simulation was utilized to illustrate a model for the adsorption of multiple DHA molecules on the BG RUs in aqueous medium. Six different systems were created consisting of BG, DHA, and H₂O molecules. For the control study, systems containing BG with only water (BG–H₂O) and BG with only DHA (BG–DHA) were considered. The details of all systems are summarized in Tables S2 and S3. Later, MD simulation was performed on the low-energy configurations (LEC) obtained from the adsorption locator analysis over 10 ns using the same force field. After that, the characteristic dynamic and structural properties such as RDF, diffusion coefficient, bond length distribution, and viscosity were investigated for all given systems.

Study of Adsorption through MC Simulation.

MC molecular simulation finds the suitable adsorption locators on the substrate molecule for the adsorption of molecules of interest.²⁹ Based on the preferable adsorbate–adsorbent interactions, it determines the possible energetically stable configurations and computes binding/adsorption energy for the resulting configurations.²⁰ Each simulation can result in multiple energetically stable adsorbate–adsorbent structures for a single system depending upon several factors such as, the number of adsorbate molecules incorporated in the simulation, variation in the spatial arrangement of the adsorbates, and differences in their resulting adsorption/desorption energies.^29,30 For instance, the BG and DHA interaction resulted in a total of 22 simulated configurations for BG-5DHA(aq), 5 for BG-7DHA(aq), 5 for BG-10DHA(aq), 23 for BG-12DHA(aq), 9 for BG-15DHA(aq), 5 for BG-20DHA(aq), 242 for BG–DHA, and 57 for BG-H₂O systems. The adsorption energies for the most stable and the least stable configurations for each system are shown in the Table 1. Among different energy parameters, the desorption energy (dEad/dNi) of DHA is crucial as it indicates how strongly DHA is adsorbed on the surface of BG. The higher the value of negative desorption energy, the harder it becomes to release the adsorbed molecule from the adsorbent’s surface and, therefore, the stronger the adsorption of DHA. The BG–DHA system resulted in −64.5 kJ/mol for DHA in its most stable configuration, while all other systems showed a comparatively low value and it decreased with increasing DHA loadings in the systems. This decrease in desorption energy indicates that the DHA molecule cannot interact strongly with BG in an aqueous system.²⁹ This seems reasonable as water and DHA are immiscible (aqueous and oil phase), and the water content is sufficiently high compared to DHA in any given system. This has further been observed when low water volume (50 H₂O count) was considered. A good number of DHAs were able to closely interact with BG, and the desorption energy (dEad/dNi) of DHA was also found to be relatively large (Table S4). This finding indicates that under low water content, DHA can effectively interact with BG, while with high water content, the affinity toward H₂O becomes more pronounced.

Table 1.

Energy (in kJ/mol) Values Computed through MC Simulation for Eight Different Systems as Shown in the Table S2^a,b

systems	total energy	adsorption energy (R + D)	rigid adsorption energy (R)	deformation energy (D)	H2O/dEad/dNi	DHA/dEad/dNi
BG–H2O–LEC	−4305.54	−4497.97	−4514.95	16.97	−0.18
BG–H2O–HEC	−4222.14	−4415.28	−4422.85	8.28	−0.13
BG–DHA–LEC	−473.88	−344.12	−270.34	−73.78		−64.54
BG–DHA–HEC	−307.63	−177.87	−92.87	−84.99		−18.34
BG–5DHA–LEC	−4515.43	−4578.1	−4514.21	−63.89	−0.15	−30.81
BG–5DHA–HEC	−4418.61	−4481.27	−4409.03	−72.45	−0.18	−30.97
BG–7DHA–LEC	−4381.71	−4392.47	−4276.99	−115.49	−0.2	−27.2
BG–7DHA–HEC	−4288.15	−4298.91	−4180.95	−117.97	−0.12	−27.81
BG–10DHA–LEC	−4477.38	−4410.29	−4232.79	−177.49	−0.15	−25.08
BG–10DHA–HEC	−4397.08	−4329.99	−4148.42	−181.56	−0.13	−19.25
BG–12DHA–LEC	−4644.52	−4525.55	−4328.94	−196.58	−0.06	−17.96
BG–12DHA–HEC	−4550.48	−4431.48	−4225.42	−206.06	−0.18	−20.62
BG–15DHA–LEC	−4712.86	−4515.98	−4265.75	−250.25	−0.06	−20.78
BG–15DHA–HEC	−4614.41	−4417.56	−4157.07	−260.49	−0.18	−20.15
BG–20DHA–LEC	−5316.64	−4990.04	−4676.61	−313.43	−0.17	−18.29
BG–20DHA–HEC	−5235.81	−4909.19	−4592.84	−316.35	−0.11	−19.46

^aThe energies for the most stable (low adsorption energy) and the least-energy (high adsorption energy) configurations are shown for each system.

^bdEad/dNi is the desorption energy required to release an adsorbed species from the substrate. Deformation energy is the energy being released after adsorption. Rigid adsorption energy is the energy for the adsorption of an unoptimized molecule on the substrate. Adsorption energy is the energy necessary for the adsorption of a flexible optimized structure on the substrate molecule. Total energy is adsorption energy− (total energy of adsorbate × total number of adsorbate molecules). LEC and HEC are low-energy and high-energy configurations, respectively.

Unlike LECs, the dEad/dNi high-energy configurations (HECs) of DHA was found to be comparatively large in all given systems. This is likely as, in a high-applied energy system, H6₂O molecules tend to set apart from the BG substrate due to low molar mass whereas DHA gets enough sites for effective interaction, which probably led to increase in its dEad/dNi. Among all six systems, the DHA desorption energy for high-energy configuration of the BG-12DHA(aq) system increased by more than 2 kJ/mol, which is probably due to its ability to better fit within the adsorption site on the BG chain within the given concentration value. Overall, the feasibility of adsorption of DHA on the BG substrate is highly influenced by the volume of water in the system. BE of the optimized LECs for eight compositions are summarized in Table S5. The variation of BE as a function of the DHA loading is shown in Figure S4, which demonstrates that the (1:12) BG/DHA system possesses the lowest energy. This also indicates that among the six systems, the adsorption of DHA on BG favors prominently in the system having 12 DHA loadings.

Figures 4 and 5 show the LECs and HECs obtained from the adsorption locator simulation for the systems consisting of both DHA and solvent molecules, while, for the control study, the LECs and HECs for the BG–H₂O and BG–DHA systems are represented in the Figure S5. From Figure 4, it is evident that H₂O can preferably adsorb on the surface of BG as, in reality, BG is a less viscous water-soluble polymer. However, for the HEC (Figure 5), the water molecules tend to move away from the BG substrate compared to DHA. This observation implies that the adsorbed DHA molecules are relatively stable and do not set apart from the surface, although the system is in its highest-energy and least stable configuration. However, for the systems containing 15 and 20 DHA molecules, the adsorption of DHA is partially hindered in both LECs and HECs and that all DHA molecules could not closely approach the BG substrate. This is likely as, for these two systems, the DHA loading per BG substrate was very high, which lowered the number of active sites for adsorption, as discussed before. This observation perfectly reflects why the desorption energy of DHA for these two systems was found critically low. It is also evident that the adsorption of DHA does not follow a definite pattern and is somewhat random on the BG substrate. However, the initial interactions seem to begin at the end sites of the BG chain and then distribute throughout the center of the RUs with increasing DHA loading. Furthermore, DHA molecules tend to crowd over the end sites more strongly than the middle region of the BG chain. This observation suggests that the end sites of the BG chain are relatively active for multiple DHA adsorption.

Figure 4.

LEC obtained through MC simulations. Figures A to F show LECs for systems containing both DHA and H₂O as the adsorbates with DHA loading being increased from 5 to 20. The adsorption of DHA seems favorable on the BG substrate for the system with 5, 7, 10, and 12 DHA loadings and starts from the end sites of the BG RUs. Furthermore, increasing DHA loading in the systems tends to lower the energy for adsorption, and thus, results in unabsorbed scattered DHA molecules in the BG-15DHA(aq) and the BG-20DHA(aq) systems.

Figure 5.

Least stable HEC obtained through MC simulation. Figures A–F show HECs for the systems shown in the previous Figure 4. In HECs, H₂O adsorbates tends to move away from the BG substrate more frequently than the DHA molecule for all systems, suggesting that the DHA molecules can strongly adsorb on the BG surface even in its least stable form.

The effects of different nonbonding interactions, namely, van der Waals and electrostatic interactions on the total energy of the systems, were also studied. The changes in the energy values as a function of iterations are shown in Figure 6 for all systems, including controls. It can be said that the nonbonding interactions were more frequent and contributed significantly to the adsorption of DHA on the BG RUs. For electrostatic interactions, the energy was found to be high for any given system, including controls. With the increase in mole fraction of DHA in the systems, the electrostatic energy component increased along the negative direction (Figure 6C–H). This implies that both DHA and H₂O molecules interacted significantly with BG utilizing electrostatic interactions. Although for van der Waals interactions, the energy contribution to the total energy of any given system was relatively small. All systems showed a slightly positive energy increase as the simulation proceeded. The computed energy components are obtained for the combined interactions of both DHA and H₂O with BG.

Figure 6.

Energy variation as a function of iteration steps for all given systems represented in Table S2. It shows the average total energy of any systems obtained from the sum of the energy contributions resulted from different nonbonding interactions, namely, van der Waals and electrostatic interactions, and intramolecular interactions. Figure A demonstrates that the electrostatic energy contributes the most to the total energy in the BG–H₂O system, while, for the adsorption of DHA, energy for both van der Waals and electrostatic interactions contribute to the total energy value (B–H).

The energy distribution curve implies the probability of a particular interaction based on energy variation. The probable energy distribution for water in the BG–H₂O system was found to be at −0.5 and −6.5 kJ/mol (Figure S6), while, for the system with DHA, the probable energy distribution was observed to be around −18, −23.8, and −29 kJ/mol (Figure S6). Therefore, water molecules show both hydrogen bonding and close-contact interactions with the BG substrate, while the interactions between DHA and BG are predominantly close-contact nonbonding interactions that become operational when DHA remains in close vicinity of the BG RUs.

RDF.

RDFs refer to the probability of finding any molecule $\text{(g(}r\text{))}$ at a distance ($r$ in angstrom) from a reference point (in this case, reference point means another molecule). In MD simulation, it indicates how the density of a particular molecule in a given fluidic system varies with respect to the molecular spacing $\left(r\right)$³¹ It is often utilized to study the characteristic short-range molecular interaction and aggregation in a fluidic system³² It can be calculated using the following eq 2

$$g(r)=\text{d}N(r)/4\mathrm{\Pi }{r}^2\rho \times \text{d}r$$ (2)

where $\rho$ refers to the local density of a molecule in a fluid, meaning the density of the molecule at a given distance $\left(r\right)$ from a reference molecule. $dN(r)$ indicates a function used for computing the number of molecules within a short distance of length $(\text{d}r)$.

To explore the details of molecular interactions between BG and DHA in an aqueous system, the changes in the RDF have been analyzed through MD simulation (Figure 7). For this observation, we studied the LECs obtained from MC simulation for all six aqueous systems with increasing DHA loading from 5 to 20. Here, the BG molecule was considered as the reference molecule with which all molecular interactions have been observed. Figure 7A shows the probability of observing molecular interactions at different molecular separations (r) between BG and DHA. For the system having five DHA loadings (1:5 BG/DHA), three peaks were observed at distances of 1.5, 2.5, and 2.8 Å, while, for the other five systems, a broadly distributed peak with centers positioned at 1.8, 2.9, and 5 Å was found. Hydrogen bonding usually occurs when molecules/atoms are 1–2 Å distance apart.³³ The molecular interaction beyond this distance is most likely to be nonbonding.^18,23 Unlike the other five systems, the RDF peak at 1.5 Å intensity obtained for the 1:5 BG/DHA system was significantly high. This could be due to the presence of a low mole fraction of DHA, as discussed earlier. Furthermore, the system with 12 DHA loadings showed a broad peak response for the highest molecular interaction. After that, the peak height gradually decreased with increasing DHA mole fraction in the systems. This observation suggests that the interacting capacity could reach the highest when 12 DHA molecules per five RUs of BG were incorporated in the system (Table S5). DHA molecules in the 1:12 BG/DHA system likely orient themselves around the BG RUs and offer superior interacting efficiency. Note that all six systems contained water molecules, which also have an affinity toward BG. Because BG is a less viscous and low-MW polymer, the hydroxyl component of BG can share hydrogen bonding with the polar solvent. Hence, both water and DHA molecules would compete over their affinity toward BG in any given system. However, due to the small size and polar nature, solvent molecules will be more prone to interact with BG and, thus, will reduce the local density of DHA around the BG RU. The inset of Figure 7A shows the RDF of the BG–H₂O system compared with BG–DHA and 1:12 BG/DHA systems.

Figure 7.

RDFs and bond length distributions for molecular interactions between BG and DHA in an aqueous system. (A) shows RDFs for BG-5DHA(aq), BG-7DHA(aq), BG-10DHA(aq), BG-12DHA(aq), BG-15DHA(aq), and BG-20DHA(aq) systems, representing possible intermolecular interaction between the BG RUs and the DHA set. It shows that the intermolecular interaction between BG and DHA is the highest for the system BG-12DHA(aq). The inset of A shows the RDFs for the BG–H₂O, waterless BG–DHA, and the BG-12DHA (Table S2) systems. (B) Effect of DHA loadings on the bond length distribution in six different systems. The inset of B shows the bond length distribution for the BG–H₂O, waterless BG–DHA (Table S2), and the BG-12DHA systems.

Additionally, the BE of the optimized LEC showed that the system with 12 DHA loading showed the lowest energy value, indicating relatively high stability of the system compared to systems with either low or high DHA loading (Figure S4). From this observation, it can be said that the 12 molecules of DHA loading per five RUs of BG ratio in a fully solvated system is the optimized ratio for obtaining the highest probable molecular interaction between BG and DHA.

Bond Length Distribution.

The distribution of bond lengths for the interactions between (1) H₂O–H₂O, (2) BG–H₂O, (3) DHA–DHA, (4) BG–DHA, and BG–DHA–H₂O was examined for all systems of interest (Figure 7B). The inset of Figure 7B shows the probable bond distribution for the H₂O–H₂O and the BG–H₂O interactions in the BG–H₂O system and the DHA–DHA and the BG–DHA interactions in the waterless BG–DHA system. Water molecules show a sharp peak at 0.96 Å, which corresponds to the H–O···H bond.³⁴ This indicates that water molecules tend to bind with strong H bonds in the BG–H₂O system. When BG and DHA were considered as the interacting species, the H–O···H peak intensity significantly reduced, which is relevant for explaining the interaction between BG and H₂O. As for BG and DHA interactions in the BG–DHA system, bond length showed peaks corresponding to $\text{O}-\text{H}\cdots \text{H}-{\text{C}}_{{\text{sp}}_3}$ (1.1 Å), C═H···H–C (1.44 Å), and $\text{C}-\text{H}\cdots {\text{C}}_{{\text{sp}}_3}$ (1.51 Å) interactions.³⁵ However, the peak intensity corresponding to 0.95 Å was significantly low, indicating very little probability of occurrence of a pure H bond in the BG–DHA system.

To understand the probable bond length distribution for the interaction of DHA with BG RUs in the presence of H₂O, the distribution plots of all six systems are represented in Figure 7B. It can be seen that the peak intensity of O–H···O (0.95 Å) and O–H···O (ring) (1.43 Å) bonds is significantly small for the system containing both DHA and H₂O, while the peak intensity of 1.1 and 1.5 Å is relatively high compared to that of the systems considered as the controls. The increase in intensity of these peaks and a decrease of 0.95 Å, respectively, depicts the better BG–DHA interactions under aqueous conditions. This also depicts that the H₂O–H₂O and BG–H₂O interactions become weak with the increase in DHA counts in the systems. Furthermore, the 1:12 BG/DHA system gave the two most intense peak responses corresponding to 1.1 $\left(\text{O}-\text{H}\cdots \text{H}-{\text{C}}_{{\text{sp}}_3}\right)$ and 1.51 Å $\left(\text{C}-\text{H}\cdots {\text{C}}_{{\text{sp}}_3}\right)$. This further indicates that all DHA molecules in this system can strongly orient themselves around BG and, thus, display the highest probability compared to other systems as observed in previous figures.

Experimental Investigation of BG–DHA Interaction Processes.

Experiments were performed with BG–DHA–water systems like the simulation setup. BG and DHA were taken at 1:5, 1:7, 1: 10, 1: 12, 1: 15, and 1: 20 loadings (Figure 8A,B). Figure 8A shows an image of the initial mixture of the BG–DHA in an aqueous system. Figure 8B shows the samples obtained after centrifuging. The brownish precipitate at the bottom is the BG–DHA colloidal particles. The images indicate that BG/DHA ratios of 1:10 and 1:12 had the most colloidal particles. Size, diffusion, and zeta potential were determined directly from DLS experiments. Experimentally obtained results were compared with the results from simulations.

Figure 8.

Comparison between experimental data and computational simulation data. (A) Experimentally prepared six formulations having 1:5, 1:7, 1:10, 1:12, 1:15, and 1:20 BG and DHA. (B) Six formulations obtained after centrifugation at 1000 rpm for 8 min. The yellow precipitate in the case of formulation with 1:12 BG/DHA indicates better BG–DHA binding. (C) DLS particle size of each formulation including BG (1). (D) Variation of zeta potential obtained for different formulations. (E) Comparison between computational (MD simulation) and experimentally found diffusion coefficients obtained through DLS size measurement. (F) Comparison between calculated viscosity values obtained for systems with different DHA loadings for both computational simulation and experimental measurements. Error bars indicate the standard deviation measurement from three experimental observations.

Size and Zeta Potential Measurement.

The BG and BG–DHA(aq) systems were prepared for size measurement according to the process described in the experimental section. Figure 8C shows the average size obtained from the DLS experiments. The size of BG-5DHA(aq) is very close to the size of BG. The reason for this observation could be the low amount of DHA present in the system. The largest particle size was obtained for BG-15DHA(aq). In BG-20DHA(aq), the size decreased slightly. It is likely because the BG-DHA system reaches equilibrium around 2.5–3.0 μm. Hence, it is likely that the BG/DHA ratio of 12 or 15 would be most effective for BG to interact with DHA for halting fat molecule digestion in the intestinal epithelial cells.

In the colloidal system, zeta potential $(\zeta )$ is the result of the interfacial interaction between static and mobile fluid planes on the surface of a particle. In this experiment, the BG in the aqueous system showed $\zeta$ around −5 mV. The $\zeta$ increased negatively until the BG-12DHA(aq) system (Figure 8D). After that, the $\zeta$ value increased in the positive direction. The reason for this might be that the 1:12 BG-DHA system has the most static DHA interface that allowed for the most negative $\zeta$ value. With increasing DHA, it could be possible that excessive DHA molecules are present on the BG surface that did not strongly interact with BG. As a result, the static and mobile interfaces did not show significant potential difference. The $\zeta$ also acts as an indicator for stability of the small colloidal particles. A $\zeta$ value less than ±5 mV would lead to rapid coagulation while that within ±10 to ±30 mV is slightly more stable. Except for BG-12DHA(aq) system, the rest show negative $\zeta$ that is < ±10 mV.³⁶ Thus, the $\zeta$ observations indicate that all the BG–DHA systems are colloidal systems, where BG and DHA are interacting primarily through electrostatic interactions and the BG-12DHA(aq) system might be the most stable among them.

Orally taken BG and fat would undergo through a series of digestion processes starting from the mouth to the stomach [a low-pH (1.1–1.2) system], which might have a negative impact on the formation of the composite. To understand the effect of stomach pH on the size of BG and BG–DHA, simulated gastric juice of pH 1.1–1.2 was used instead of water in making the BG–DHA system. Figure S7A shows that the dissolution of BG in gastric juice has no effect on the size of the BG. In both neutral water and gastric juice, the size of the BG was found to be almost similar. This finding suggests that BG fibers can withstand a highly acidic environment in the stomach and do not undergo depolymerization.¹⁴ However, when the BG–DHA conjugated system was prepared in gastric juice, it rather breaks up the DHA molecule and helps associating it with BG more effectively. The inset image of Figure S8B of the conjugated BG–DHA system in gastric fluid confirms that due to prominent incorporation of DHA with BG, the color of the sample turned more brownish. Furthermore, when compared with the aqueous (1:10) BG–DHA system, the size of the BG–DHA obtained from gastric fluid was observed to be low (Figure S7B). This confirms that although BG can resist structural alteration in a high-pH environment, the DHA molecule cannot resist the effect and rather breaks down into small compartments and binds with BG.

The zeta potential measurement of the BG and BG–DHA systems prepared in gastric fluid showed less negative surface potentials when compared with the previous findings obtained in neutral water (Figure S7B). This low zeta potential is likely observed due to high proton content (low pH condition) in the medium.

Comparison of Experimental and Simulated Diffusion and Viscosity.

The diffusion and viscosity measurements were carried out for the simulated and experimentally prepared BG–DHA systems. Figure 8E,F shows the comparison plots for diffusion and viscosity of experimental and simulated BG–DHA systems. The diffusion coefficient (D) of all six systems was analyzed through MD simulation at a constant temperature (300 K) and volume. The value of D was calculated by using the slope of the MSD versus time (ps) plot.³⁷

$$D=1/6N\underset{t\to \infty }{\lim }\frac{d}{\text{d}t}\sum _{i=1}^N({\left[r_i(t)-r_i\left(t_0\right)\right]}^2)$$ (3)

Here

$$\mathrm{MSD}(\mathrm{\Delta }t)=\sum _{i=1}^N({\left[r_i(t)-r_i\left(t_0\right)\right]}^2)$$ (4)

$r$ is the displacement from zero time point $\left(t_0\right)$ to the final simulation time $(t)$. $N$ is the number of atoms diffused in the system.

The MSD depicts the intrinsic mobility based on the molecular vibration of the molecules of interest (BG, DHA, and H₂O) in a system. The variation of MSD as a function of simulation duration is shown in the Figure S8, demonstrating an increase in the system’s degree of dynamic mobility over simulation time. Throughout the dynamic simulation trajectory, the DHA molecules tend to remain close to the BG polymer even though the system was in constant movement.

Figure 8E shows the plot of the diffusion coefficient as a function of the number of DHA molecules loaded in different systems. The diffusivity of the DHA molecules toward the BG substrate was found to be the highest at a low number of DHA in both the simulated and experimental observations. The simulated and experimental D values were 6.95 × 10⁻⁹ and 6.91 × 10⁻⁹ m²/s, respectively. As the DHA loading increased, the rate of diffusion attenuated proportionally with D for both the experimental and simulated systems. However, the amount of attenuation for the experimental and simulated systems was different. The simulated data showed a greater decrease in diffusion compared to the experimental systems. For 20 DHA loading, $D$ was calculated to be 9.17 × 10⁻¹⁰ and 2.37 × 10⁻⁹ m²/s, respectively, for simulated and experimental systems. The higher $D$ value in the experimental observation was likely due to the low-energy interaction of BG–DHA with water.¹⁸ This observation implies that the system attains a viscous flow state upon increasing DHA concentration.

To better understand the change in viscosity with DHA loading, viscosity analysis was carried out for both the simulated and experimental data using the Stokes-Einstein eq 5³⁸

$$D=\frac{k_{\text{B}}T}{6\text{π}\mathrm{\eta }{r}_{\text{o}}}$$ (5)

Here, $k_{\text{B}}$ is the Boltzmann constant and $D$ is the diffusion coefficient of $r_{\text{o}}$ radius particle in a solution with $\eta$ viscosity at $T$ temperature. Similar to the $D$ values, the viscosity of lower-DHA loading systems is in good agreement. As the DHA loading increased, the viscosity values deviated significantly for simulated and experimental systems. In both the experimental and simulated data, the BG-7DHA system showed a sudden increase in viscosity. The reason for this increase is not apparent. After that, the viscosity steadily increased for both systems. However, the simulated system showed much higher viscosity compared to the experimental systems. The likely reason for the failure of high-DHA loading simulations is to account for the low-energy interactions.¹⁸ However, in the case of both diffusion and viscosity analysis, the experimental and simulated systems showed excellent agreement when compared for low-DHA loading systems. This indicates that the simulations were able to accurately predict the interaction and energy values for BG–DHA systems. Thus, these molecular analyses can be reliably used to predict the behavior of BG and DHA in aqueous systems.

In Vitro Study of Permeability of BG–DHA through Intestine.

In this section, we performed an in vitro intestinal permeation assay of DHA at seven different time points (30 min, 1, 2, 3, 4, 5, and 6 h, respectively) to study whether BG fibers can reduce DHA diffusion through the intestine. To track the amount of DHA diffusing through the intestine, we used RhB-conjugated DHA (DHA–RhB) and BG–DHA–RhB systems. In this study, the porcine intestine was utilized upon which DHA–RhB and BG–DHA–RhB were loaded. A fully automated flow-through system associated with a multi-channel pump was used to ensure continuous flow of PBS at a flow rate of 4 mL/h below the intestine. The PBS passing below the intestine catches the diffused formulations at different time points that were initially loaded on top of the intestinal blocks and was collected through an outlet channel. The collected samples were subjected to endpoint absorbance analysis at a wavelength of 550 nm. Figure S9A shows the plot of absorbance versus time plot that demonstrates the amount of BG–DHA–RhB that diffused through the intestine at different time points compared to DHA–RhB in terms of absorbance intensity of RhB. From the figure, it can be seen that the value of absorbance of RhB for BG–DHA–RhB loading was significantly low compared to DHA–RhB, which indicates that the diffusion of DHA through the intestine reduced prominently when conjugated with BG fibers. The maximum concentration of diffused DHA was obtained at 5 h time period and was calculated to be 0.48 μg/μL. This value is significantly lower than the maximum DHA concentration obtained for DHA–RhB loading. This analysis suggests that when DHA binds with BG, it is no longer capable to pass through the intestine. The value of concentration of diffused DHA was determined from the standard curve represented in the Figure S9B.

DISCUSSION

In this study, we aimed to investigate how dietary BG fibers help in delaying lipid digestion in the small intestine through computational and experimental observations. In order to do this, a DHA fat molecule, best known as fish oil, has been chosen to see whether it shows facile interaction with the BG fibers. Once ingested, BG reaches the stomach and is dissolved in highly acidic aqueous medium of the gastric juice. However, the pH of the fluid is slightly basic in the small intestine, which is why all molecular interactions were analyzed in neutral aqueous medium. It has been observed under aqueous conditions, the probable molecular interactions between the BG polymer and the DHA molecule become more feasible. For instance, as from the bond length distribution and RDF analysis, it was found that both H bonding and nonbonding (electrostatic and van der Waals) interactions were involved, which allowed better intractability and association of DHA with the BG molecules. However, the probability of interaction is highly dependent on the mole ratio of DHA/H₂O in the system. Another key finding of this study is the estimation of the diffusion coefficient value for different systems from which the viscosity of each system is predicted. As per analysis, we observed that the viscosity of the systems proportionally increased (or decreased D value) with increasing DHA mole fraction. The viscosity of the BG fibers is one of the critical factors involved in the fat trapping mechanism. The mechanism involves the formulation of a viscous colloidal system with fat molecules that delays further lipid digestion and subsequently inhibits the absorption of the nutrients through the intestine.¹⁰ It has been proposed that BG, due to its gel-forming ability, can potentially prevent the recirculation of the bile acids (formed through conversion of cholesterol in the liver) through the small intestine.¹⁴ Both experimental and simulation findings strongly correlate with this idea and interpret the interactions that lead to the formation of a geltype colloidal state between the BG fibers and the DHA molecules up to a certain loading amount of DHA. The effect of the aqueous system in this formulation is prominent. As seen from MD simulation, the intractability of BG fibers with DHA occurred via H bonds and nonbonding interactions. It is likely that the water molecules, due to their highly polar nature (higher electronegativity), functions as the oil in the water emulsion system in the presence of insoluble fat substance and therefore leads to facile interactions between the hydrophobic compartments of BG and DHA. Whereas BG fibers, due to their extended organic framework, show substantial fat associating efficiency in polar solvent medium. This hypothesis has been previously studied by Zhai et al., who showed that the effect of BG fibers on lipid emulsification and subsequent depletion of lipolysis depends on the MW, mixing methods, the structure and size of lipid droplets, and the oil emulsion to a great extent.¹⁶ Zhi et al. reported that the structural orientation that the BG fibers adopts to emulsify lipid droplets is highly controlled by the medium in which they interact. Note that if the concentration of the BG dietary fibers is increased in the system, it would lead to depletion flocculation, meaning that the corresponding depletion force would be large enough to overcome the repulsive energy barrier that holds droplets apart.²⁰ Furthermore, changes in the solvent loading can significantly influence the findings such as optimum micelle formation for a specific system. However, altering the type of interactions and molecular intractability of BG with DHA is unlikely as observed from this study and the diffusivity and viscosity properties of BG fibers on the DHA emulsification. The in vitro intestinal permeation assay also confirmed that the effect of BG on the diffusion of DHA through the intestine is inevitable. DHA becomes significantly ineffective when binds with BG. Overall, these findings can help us in understanding more and elucidate the type of interactions between the BG polysaccharide and DHA at the molecular/atomic level that played a role in delaying lipid digestion and, eventually, lowering the lipid absorption in the GI tract. The mechanism of how BG plays a critical role in delaying fat digestion in the small intestine is illustrated hypothetically in Scheme 1.

Scheme 1.

Schematic Illustration of How BG Functions to form a Colloidal Viscous State with the Large Fat Molecules in the Small Intestine Based on the Interaction Mechanism Proposed in This Study; (A) Usual Pathway for Fat Molecules to Be Absorbed in the epithelium Cells of the Small intestine; (B) Interaction of BG and Fat Molecules Would form a Colloidal System that Halts the Digestion and Absorption; and (C) Scheme Showing Fat Molecule Absorption (a) and Absorption Halting through Interaction with BG (b) at the Epithelium Cells of Microvilli

CONCLUSIONS

In this study, we investigated molecular interaction between BG and DHA through both experimental and computational modeling and studied how aqueous medium affects the physiological properties of BG for their facile interaction. The DFT simulation results of structural optimization, IR analysis, COSMO profiles, PDOS, and Mulliken charge distributions showed that BG and DHA interact with each other through nonbonding forces in aqueous medium. Force field theory-based MCSs showed that BG–DHA interacts through hydrogen bonding and electrostatic forces. Also, even though the contribution from van der Waals interaction is mostly quenched by the water molecules, there is still a small contribution toward stabilizing the BG–DHA systems. The MDSs showed that the BG–DHA forms stable systems in aqueous medium over 10 ns of simulations. Experimental observations revealed the changes in size and zeta potential of BG–DHA colloidal particles with DHA loading. Furthermore, the diffusion and viscosity analysis for both simulated and experimental observations were in excellent agreement, indicating the feasibility of the simulation system. Based on these results, we propose that BG and DHA interact through nonboning forces in the aqueous system that produces a viscous colloidal state, which could be the potential mechanism for fat trapping and subsequent delay in lipid digestion in the GI tract. The key findings of this work are (1) that BG–DHA interaction is the highest when their ratio is 1:12 and (2) that, of the nonbonding forces, the electrostatic force is the one that contributes the most in the BG–DHA interaction process. Hence, tuning the electrostatic intractability of BG is likely to increase its fat capturing capability as observed from the in vitro diffusion study. This would further guide us in formulating different BG fiber-based nano-therapeutic carriers for obesity treatment.