Influence of Amidation on the Release Profiles of Insulin Drug from Chitosan-Based Matrices

Abstract

The present study deals with the comparative analysis of insulin drug release from pure chitosan (CS) and its crosslinked amide derivatives. The objective of this study was to investigate the influence of fatty acid derivatives on the release profiles of insulin drug from CS-based matrices. In order to form cross-linked CS-based amide derivatives, the CS biopolymer was reacted with four different fatty acids with varying amount of unsaturation, including stearic acid (SA), oleic acid (OA), linoleic acid (LA), and linolenic acid (LLA), and then subjected to cross-linking. Following this, the pure CS and cross-linked CS amide derivatives were loaded with insulin drug and were characterized thoroughly by making use of various instrumental techniques such as FTIR, UV–Vis, TGA, HRTEM, DLS, PDI, and zeta potential studies. In addition, the insulin release profiles were studied and compared between pure CS and CS amides at two different pHs, 7.4 and 1.2. Finally, the insulin drug release data was investigated with five different pharmacokinetic models (zero, first, Higuchi, Kersmeyer–Peppas, and Hixson models). From the analysis, the cross-linked CS amides was found to be superior to pure CS, and within the amide derivatives, the one with a high amount of unsaturation, LLA-derived CS biopolymer, was shown to be better for the release of insulin drug by means of the diffusion and dissolution pathways.

1. Introduction

Diabetes mellitus (DM) is one of the commonly found diseases in humans worldwide by its high occurrence and secondary effects. Global diabetes prevalence in 2021 was estimated to be 10.5% (536.6 million people), rising to 12.2% (783.2 million) in 2045 [1]. This stage of diabetes occurs when the human body is not able to produce or use insulin (generated inside the body) effectively, leading to high blood glucose levels, known as hyperglycemia. There are two types of diabetes, type 1 and type 2. Type 1 is caused by an autoimmune reaction of β-cells in the pancreas, and type 2 is caused by dysfunctional or defective pancreatic beta cells, resulting in decreased insulin signaling and pancreatic secretion. Insulin is used for the treatment of type 1 diabetes and advanced stages of type 2 [2]. Insulin is a polypeptide hormone normally made by the pancreatic β-cells [3]. The structure of insulin is made up of a dimer of two peptide chains linked with two disulfide bonds, where one chain is composed of 21 and other chain contains 30 amino acids [4]. The drug insulin is regularly administrated through injection due to bioavailability issues, but this causes pain, local tissue necrosis, bacterial contamination, and nerve damage by repeated subcutaneous injections. Alternatively, the oral route is available, but suffers from the weak stability in the gastrointestinal tract under varied pHs, sensitivity to enzymatic degradation, hydrophilicity, and low bioavailability due to its large size, which leads to lower intestinal absorption [2,5,6]. Hence, due to its inconvenience of conventional treatment (injection or oral), and to improve the quality of lives, alternative methods of insulin delivery have been extensively explored [7]. In this view, a successful oral protein nanocarrier that is self-stable in the gastrointestinal tract, high drug absorption, site-specific release, and mucoadhesive properties have been proposed for the delivery of insulin drug [6]. Therefore, several researchers have carried out studies of insulin encapsulation into polymer NPs as a good strategy to improve its oral bioavailability. This is based on the fact that the polymer capsule protects its loaded insulin from enzymatic degradation in the acidic environment of stomach, and at the same time, the nanosized carrier system enhances the absorption. Of the many different polymers used as nanocarrier systems, the studies based on chitosan (CS) biopolymer have demonstrated beneficial activities in terms of protecting and proliferating the pancreatic β-cells, lowering hyperglycemia, and preventing impaired lipid metabolism associated with diabetes mellitus [8,9,10,11].

CS is a natural and widely available cationic polysaccharide, derived from chitin (found in the shells of shrimps, crabs, and lobsters), which is considered the second most abundant natural biopolymer after cellulose [8]. CS consists of β-(1-4)-linked D-glucosamine (deacetylated unit) and N-acetyl-D-glucosamine (acetylated unit), and due to its unique structure and associated functionality, this biopolymer has applications in many different sectors, including environmental engineering, textile industry, papermaking, food industry, cosmetics, medicine, biotechnology, and pharmaceuticals. In addition, a sudden increase of the utilization of CS, and its composites in particular, in the biomedical sector during recent years has been due to its antimicrobial and antifungal properties, and its ability to form biofilms, hydrogels, and nanoparticles (NPs) with no compromise in biocompatibility. CS is well suited to the human body, due to its biocompatibility, biodegradability, mucosal tolerance, and non-toxic properties. However, the applications of CS are limited in spite of its very attractive physicochemical characteristics, and the main issue with this biopolymer is its poor water solubility, as it dissolves in diluted/weak acid solutions, such as acetic acid [12,13]. Alternatively, the chemical structure of CS, with its large number of amino (–NH₂) and hydroxyl (–OH) functional groups, can be modified to obtain derivatives of CS with much improved solubility, stability, and mechanical and thermal resistances. For example, the protonation of amine groups of CS under acidic conditions generate positive charges (–NH₃⁺), which further interacts with the negatively charged mucosa membranes and gets adsorbed easily on the mucosal surface, thereby improving the adhesion rate, antimicrobial activity, and cell membrane permeability [14]. Taking into consideration this physiological point in view, CS biopolymer when used as a drug carrier has the special characteristics of mucosal adhesion, increased drug stability, enhanced residence time, controlled administration route, increased targeting ability, controlled drug release, improved dissolution of drugs with poor solubility, and ability to adjust the tissue permeability of hydrophobic drugs [15].

CS and its derivatives, like any other material, have strongly influenced physicochemical properties due to alterations in chemical composition and associated molecular sizes. In this view, unmodified CS has poor water solubility at physiological pHs, which results in low drug delivery efficiency and cellular uptake due to a very high molecular weight and viscous nature. As compared to several other polysaccharides, CS easily allows several modifications to its basic structure. One approach to introduce new functional groups to the backbone of CS is by grafting [16]. The amino group of CS is a highly reactive functional group, and its modification allows for the formation of derivatives with improved water solubility and sophisticated properties for biological applications. In addition, for the delivery of low-molecular-weight drugs, peptides, and genes through CS-based drug delivery systems, the efficient way of utilizing CS biopolymer is by removing the hydrogen atoms from free amino groups and forming the amides through various means [17]. In addition, the amides formed by grafting hydrophilic CS chains with hydrophobic compounds like carboxylic or fatty acids produces products of amphipathic behavior, which are able to self-assemble and generates NPs under suitable conditions [16].

The usage of fatty acid materials for the grafting of CS biopolymer offers several chemical and biochemical advantages, for example, the coating stops the polymer from degrading and protects it in the acidic environment of stomach. The fatty acids serve as important components of body cells, and hence bonding/coating with fatty acid molecules helps to overcome the biopolymer’s barrier properties, such as water solubility, physiological pHs, and bodily enzymes, all of which are factors that contribute to the lessening of total drug efficiency. Therefore, considering the advantages of altering the functionality of CS moieties using the fatty acid molecules, the present study is aimed at developing an effective insulin drug carrier system. For that, we have selected four different fatty acids for the modification of chemical functionality of CS: oleic acid (OA), linoleic acid (LA), linolenic acid (LLA), and stearic acid (SA). The selected acids OA, LA, and LLA are long-chain unsaturated fatty acids, while the SA is saturated, but they all share the same number of carbons atoms (C₁₈). It has been hypothesized that the drug-carrying polymer (modified with a combination of animal and plant fat through saponification and acidification), when inserted into the phospholipid bilayer of the cell membrane, creates disorder and thus leads to increased permeability [18,19]. In addition, the fatty acid groups with their hydrophobic part enhance the stability of insulin drug under all kinds of biological environments.

2. Experiment

2.1. Materials and Methods

All chemicals were used as they were received and without any further purification. The high-molecular-weight chitosan (CS), the four fatty acids (oleic acid (OA; 65–88%), linoleic acid (LA; ≥99%), linolenic acid (LLA; ≥99%), and stearic acid (SA; 95%)), methanol (99.8%), and dichloromethane (99.9%) were purchased from Sigma-Aldrich (Jinan City, Shandong Province, China). The other chemicals, such as acetic acid (99%), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) used as catalyst, and the crosslinker sodium tripolyphosphate (TPP), were obtained from Loba Chemie (Mumbai, India). Insulin aspart from NOvoRapid^® FlexPen (Bagsværd, Denmark), with each mL containing 100 units of insulin drug, was used for the loading. A dialysis membrane (MWCO 3500) was obtained from Thermo Scientific (Rockford, IL, USA), and deionized water was used across all the treatments.

2.2. Synthesis of CS-Crosslinked Fatty Acid Amides

For the formation of crosslinked CS–fatty acid amides, the first step involves the formation of an amide bond between CS and fatty acids such as OA, LA, LLA, and SA through an EDC–mediated reaction. To form an amide bond, about 1 g of CS was dissolved in 100 mL of 1% aqueous acetic acid solution, followed by the addition of 85 mL of methanol. Then, the fatty acid (OA, LA, LLA, and SA) was added to the CS solution (0.34 mol/mol glucosamine), followed by the dropwise addition of 15 mL of EDC-methanol (0.07 g/L) solution with constant stirring at room temperature. For this reaction, the mole ratio of fatty acid to EDC was maintained at 1, and after 24 h of stirring, the reaction mixture was transferred to 200 mL methanol/ammonia solution (7/3, v/v) with stirring. The precipitate of CS-g-fatty acid amide hydrogel obtained was filtered, washed with distilled water–methanol mixture, and dried at 20 °C under vacuum for 48 h to obtain the powder [20].

In the second step, which involves the crosslinking reaction, about 0.25 g of the obtained CS-g-fatty acid amide hydrogel was dissolved in 50 mL of 0.1 M acetic acid solution, followed by the addition of methylene chloride (5%, v/v) and homogenization for 2 min at 12,000 rpm. By making use of a rotary evaporator under vacuum at 450 atm for 30 min at 20 °C, methylene chloride solvent was separated, and after that, the TPP solution (10 mL, 8% w/v) was added with constant stirring at room temperature. The mixture was allowed to stir for an additional 30 min, and then it was filtered, the precipitate was washed, and it was dried in a vacuum for 48 h to finally obtain the crosslinked product of CS-g-fatty acid amide. The schematic representation of the formation of crosslinked CS-fatty acid amides is provided in Scheme 1.

2.3. Studies of Insulin Drug Loading and Release

Insulin loading was accomplished by immersing 0.5 mg of CS-g-fatty acid amide hydrogel in 1 mL (containing 100 IU (international unit)) insulin solution for 24 h under stirring at room temperature. After this period, the suspension was separated by filtration, and washed with distilled water. The free and unbonded insulin was measured in the supernatant by making use of the UV–Vis spectrophotometer (Perkin Elmer Lambda 45, American Laboratory Trading (ALT), East Lyme, CT, USA) at a wavelength of 284 nm, where the encapsulation efficiency (EE) was measured using the following equation:

EE (%) = [(Total amount of insulin) − (Free insulin in the supernatant)]/(Total
amount of insulin) × 100

To study the release of insulin from the crosslinked CS-g-fatty acid matrices, two different solution media were used in a dialysis membrane. One solution involves synthetic stomach fluid (SSF) with a pH of 1.2 and was prepared using distilled water added with HCl (0.42 M) and glycine (0.4 M) [21]. The second medium is the simulated intestinal fluid (SIF) that was prepared by dissolving one buffer tablet of pH 9 in 200 mL distilled water, followed by the dropwise addition of 1 M HCl to adjust to a pH of 7.4.

For the release, about 0.5 mg each of the insulin-loaded crosslinked CS-g-fatty acid matrix was placed in a dialysis bag (7 cm length and 22 mm diameter; MWCO 3500) containing 2 mL of either SSF (pH 1.2) or SIF (pH 7.4) buffer solution. The dialysis bag with the drug composite was immersed in 7 mL receiver solution (SSF or SIF) identical to the inner solution, followed by incubation in a water bath at 37 °C with stirring. An aliquot (250 µL) of the receiver solution was collected at specified time intervals (1, 2, 3, 4, 5, 6, 12, 22, and 24 h) and replaced with an equal amount of the same fresh medium. The amount of insulin and its concentration available in the aliquots were determined by using the UV–Vis spectrophotometer (Perkin Elmer Lambda 45, American Laboratory Trading (ALT), East Lyme, CT, USA) at 284 nm wavelength [22].

2.4. Instrumental Analysis

The functionality, bonding, and chemical structure of as-synthesized CS hydrogels before and after the amidation, crosslinking, and insulin drug loading were analyzed using Fourier transform infrared spectroscopy (FTIR) (Nexus 6700 FTIR; Nicolet Magna, Jersey, NJ, USA). The samples were prepared by the KBr pellet method, and the analysis was run in the spectral range of 400 to 4000 cm⁻¹. In general, for the UV–Vis spectroscopic analysis, a Perkin Elmer Lambda 45 spectrometer (American Laboratory Trading (ALT), East Lyme, CT, USA) with a wavelength set at 284 nm for insulin drug was used. The dynamic light scattering (DLS) studies were used to determine the particle size, polydispersity index (PI), and zeta potentials of as-synthesized CS-based hydrogel particles. For the sampling, the dried hydrogel powders were ground and then dispersed in distilled water for analysis using Nanoplus Zeta Potential and Nano Particle Analyzer (Malvern Instruments, Malvern, UK). The thermal stability of the hydrogel samples was tested using thermo-gravimetric analysis (TGA) on a PerkinElmer TGA-7 Thermo-gravimetric Analyzer (American Laboratory Trading (ALT), East Lyme, CT, USA) at a heating rate of 10 °C/min, starting from room temperature and going up to 800 °C. For the investigation of morphology and orientation, transmission electron microscopy (TEM) studies on JEM 1400 Plus–HC FC (JEOL Ltd., Tokyo, Japan) was used, where the samples were prepared by thoroughly dispersing the hydrogels in distilled water and by drop-casting a sample drop onto a carbon-coated copper grid, followed by drying.

3. Results and Discussion

3.1. Physicochemical Analysis

In the present study, the biopolymeric and drug releasing properties of CS matrices were studied by grafting the CS hydrogel with different fatty acids (LA, OA, LLA, and SA) to form an amide bond, followed by its crosslinking in the presence of TPP. Figure 1 shows the comparison of FTIR spectrums of various CS-based fatty acid amide matrices before (A) and after (B) crosslinking with TPP. In this, the FTIR analysis specifically identifies the presence of amide groups and phosphate groups as a means of confirmation for the occurrence of amidation and crosslinking reactions. The comparison of pure CS spectrum and CS-amide (Figure 1A) showed some common absorption characteristic zones, that is, the peak at 3428 cm⁻¹ corresponds to the O–H overlapping the N–H stretching of primary amine, 2925 cm⁻¹ is for the C–H stretching vibration of CH₂, and 1655 cm⁻¹ is attributed to the N-H bending. In addition, in the CS-modified samples, the vibrational bands in the range of 1568 to 1744 cm⁻¹ are from the C=O of carbonyl amide groups, and the other peak observed around 1081 cm⁻¹ is attributed to the C–O groups stretching. In case of crosslinked CS (Figure 1B), the O-H and N-H bands become sharper, meaning that the hydrogen bonding is being enhanced, and the stretching vibrational band of the carbonyl group (–C=O) has shifted a bit due to the interaction between NH₃⁺ groups of CS and phosphate groups of TPP. Furthermore, the P=O bond peak for all the TPP crosslinked CS samples is observed around 1100–1170 cm⁻¹.

The comparison of UV–Vis spectroscopic analysis of pure and fatty-acid-modified CS samples loaded with insulin drug are provided in Figure 2a–e. Insulin has the maximum absorption of the 284 nm wavelength, and for the drug-loaded samples, an increase in the absorbance in the wavelength range of 280–290 nm confirms the successful loading of insulin drug within the matrices of pure CS, and also the fatty-acid-modified ones.

Figure 3 shows the TGA of pure CS compared with that of CS–fatty acid amides in the temperature range from room temperature to 800 °C. As observed from the trend of the graphs, weight loss can be noted from the beginning and the initial weight loss up to 150 °C is due to the removal of physically adsorbed water. The amount of adsorbed water was decreased in the grafted samples compared with the pure CS due to the higher hydrophobicity of the former as a result of grafting using in OA, LA, LLA, and SA, which contain long alkyl chains. The thermal degradation of the prepared materials began at about 270 °C and almost ended at about 500 °C. Moreover, it is clear that the grafting improved the thermal stability, as grafted CS has a higher thermal stability than the pure one. Additionally, the high weight retention of pure CS, at 600 °C, can be attributed to the intra- and intermolecular hydrogen bonding in CS, which decreased with grafting and led to a great decrease in the weight retention in the grafted samples.

The morphological analysis of pure CS compared with that of all other CS-based samples followed by loading with insulin drug is provided in Figure 4a–e. From the microscopic images, it can be noticed that the film-like morphology of pure CS (Figure 4a) becomes thick and develops more of an agglomerated nature due to the cross-linking in the presence of TPP (Figure 4b–f). In addition, one can observe a clear difference in the morphological characteristics of cross-linked CS (Figure 4b) with that of other grafted CS–amide composites (Figure 4c–f). This indicates that the amidation and cross-linking of the CS polymer has a strong influence on the porosity, and among the various CS–amide samples, the one with a high amount of unsaturation appeared to be more densely packed compared to the other derivatives.

The particle diameter (in solution phase), polydispersity index (PDI), and zeta potential as determined by the DLS measurements are provided in Figure 5 and the data in

Table 1. Comparison of the particles size, PDI, and zeta potential data for various CS-based samples.

Sample	Particle Size (nm)	PDI	Zeta Potential (mV)
Insulin	---	---	0.64
Pure CS	1043	0.61	9.60
CS-g-LA	2380	1.23	32.1
CS-g-OA	2769	1.41	12.22
CS-g-LLA	1071	0.62	7.83
CS-g-SA	1227	0.68	1.30

. The particle size analysis determines the grain diameter in the solution phase, while the PDI describes the homogeneity of particles and their distribution. A lower value of PDI indicates a more uniform size of particles, and therefore the particles distribution is more monodispersed. From the analysis, the sizes of pure CS and FA-modified CS particles are in the range of ~1000 to 2000 nm, that is, the particle diameters in solution are almost close to that of pure CS (1043 nm), CS-g-LLA (1071 nm), and CS-g-SA (1227 nm). However, the size is doubled in the case of CS-g-LA (2380 nm) and CS-g-OA (2769 nm) samples, and this could be due to the high amount of water absorption. In addition, the data indicates that the particle diameters in DLS measurements have significantly increased compared to the powdered samples analyzed by the TEM micrographs, meaning that the swelling nature of synthesized particles in water increases their size from nanoscale to microscale. Further, similar to the particles sizes, the PDI and zeta potential also follow the same trend, as the measured PDI of pure CS (0.61), CS-g-LLA (0.62), and CS-g-SA (0.68) samples were found to be close to each other, while they doubled for CS-g-LA (1.23) and CS-g-OA (1.41). Such an observation of significant variation in the PDI of CS-g-LA and CS-g-OA samples indicates the higher uniformity and highly monodispersed nature of the other CS samples (pure CS, CS-g-LLA, and CS-g-SA). In a similar way, the measured zeta potentials values are 9.6 mV for pure CS, 7.83 mV for CS-g-LLA, 1.3 mV for CS-g-SA, 32.1 mV for CS-g-LA, and 12.2 mV for CS-g-OA. This high zeta potential value provides information about the larger-sized hydrodynamic particles (for CS-g-LA and CS-g-OA samples), where the particles become irregularly shaped and free from their natural monodispersed nature. The zeta potential measures and quantifies the surface charge of NPs in colloidal solutions. The CS particles, with their positive surface charges (supported by NH₄⁺ groups), attract a thin layer of opposite charge and bind to it, and in this way, the insulin drug with the COO⁻ groups are attracted and form strong electrostatic bonds [16]. After the drug loading, we obtained an overall positive zeta potential for all the particles, which is a good sign for the occurrence of mucoadhesion, that is, the positively charged particles have a high amount of interaction with the negatively charged sialic groups of mucin [23].

3.2. Insulin Drug Loading and Release Studies

In this study, the drug nanocarriers were successfully synthesized by ionic gelation method using biodegradable CS crosslinked with TPP. The TPP was chosen rather than other cross-linkers because of the sensitivity and low toxicity, and because there is no possibility of causing antigenicity. Briefly, the multivalent anions (–P₃O₁₀⁵⁻) interact with –NH₃⁺ (after CS has been protonated under acid condition using acetic acid) by inter- and intramolecular cross-linking interaction, serving as the basis of the ionic gelation process for the formation of CS NPs [24].

The drug is loaded between the hydrogel’s cross-linked matrix without any chemical reaction, and the release occurs by diffusion, as over time insulin flows from the polymer matrix (high concentration) to the target site, that is, the cell (low concentration). For this phenomenon, there has to be an interface between the drug and cell fluid, as the carrier material is naturally available biodegrading polymer, and so the ridding occurs naturally. The schematic representation of the drug releasing mechanism from the CS-based hydrogel matrices has been provided in Scheme 2.

Table 2. Comparison of encapsulation efficiency (%), equilibrium time (h), and release at two different pHs for the insulin drug.

Sample	Encapsulation Efficiency (%)	Equilibrium Time (h)	Drug Release (%)
			pH 7.4	pH 1.2
Pure CS	53.09	12	35.39	99.8
CS-g-LA	55.96	24	80.7	55.4
CS-g-OA	74.52	22	59.27	26.6
CS-g-LLA	83.38	24	86.7	54.2
CS-g-SA	71.65	22	58.9	17.5

represents the insulin drug entrapment of various CS-based hydrogel matrices. From the table, it can be seen that the insulin encapsulation efficiency (%) is comparatively high for the fatty-acid-modified CS samples compared to that of pure CS. Among the four CS-modified samples, a high amount of insulin loading was observed for the CS-g-LLA sample (83.3%), followed by CS-g-OA (74.5%), CS-g-SA (71.6%), and CS-g-LA (55.9%). It may be a very difficult task to explain such a difference in the loading of the drug, as the four selected derivatives maintain the same number of carbons, and the variation is only the amount of unsaturation. Considering the fact that LLA derivative has three unsaturated bonds, the drug loading is exclusively high, while the LA derivative has very low drug loading and is even lesser than the SA derivative (a carbon chain with no unsaturated double bonds). This provides the information that the insulin loading onto the CS polymer and its derivatives is not just through a chemical means of bonding, but also through its being encapsulated through the porous kind of architectures provided by the CS base.

Further, Figure 6 provides the comparison of insulin drug release profiles carried in simulated pH conditions of gastric and intestinal fluid for pure CS and fatty-acid-modified CS particles. From the release profiles, it can be observed that the pure CS (without amide linkage) excreted almost all the drug in acidic medium within 24 h, while only 30% is released at pH 7.4, and this can be linked to the faster degradation/decomposition under acidic medium and the very much conserved behavior in the neutral mediums. However, at the same pH, the fatty-acid-modified CS samples have better release behavior, and the highest was observed for CS-g-LLA (86.7%), followed by CS-g-LA (80.7%), and it was almost same for CS-g-OA (59.2) and CS-g-SA (58.9). At a pH of 1.2 as well, the fatty-acid-modified CS samples behave in a controlled way, where high and similar drug release profiles were noted for the CS-g-LA (55.4%) and CS-g-LLA (54.2%) samples. Although this release is less compared to the pure CS sample, it indicates the stability, resistance, and controlled release behavior of the carrier under strong acidic conditions. The other samples of CS-g-OA (26.6%) and CS-g-SA (17.5%) show low release within 24 h in the same acidic medium and are considered to have strongly protected the loaded insulin drug. Such an observation indicates the importance of forming an amide bond on the CS biopolymer so as to generate a hydrophobic layer of amino acid, which serves as a coating material and protects the free/uncontrolled release of insulin drug from its decomposition in the stomach. Additionally, the equilibrium time achieved for pure CS was very low, 12 h, compared to the fatty-acid-modified CS matrices, which have equilibrium times in the range of 22–24 h. This also stresses the need for the incorporation of more unsaturated fatty acid groups onto the CS polymer in order to enhance the insulin drug loading and for controlled release over prolonged time periods. The results from our analysis follow the trend of enhanced equilibrium time with increased unsaturation, and thereby support our hypothesis that incorporating more unsaturation in the polymer chains improves the carrier’s drug loading capacity and associated equilibrium time. Therefore, from the drug encapsulation and release profile studies, the fatty-acid-modified CS polymers are shown to be better performing carriers than pure CS, and of the four modified samples, stability, resistivity, encapsulation efficiency, and controlled release nature were all shown for the sample with a high amount of unsaturation, that is, CS-g-LLA.

3.3. Drug Release Kinetics

To understand the kinetics of insulin release from various CS-derived matrices (CS-g-LLA, CS-g-LA, CS-g-OA, and CS-g-SA) due to fatty-acid-mediated amidation reaction, the drug-loaded composites were studied with different pharmacokinetic models and compared with pure CS release (without amidation). The studied pharmacokinetic models include the zero order, first order, Higuchi model, Korsmeyers–Peppas model, and Hixson model. Since the basis and formulas used for the analysis of the five pharmacokinetic models have been discussed elaborately in our earlier publications, we will not explain the same equations here, which were used to obtain the graphical parameters [19,20]. The graphs provided in Figure 7, and also in the Supporting Information of Figures S1–S5, indicate the five pharmacokinetic models at two different pHs of 7.4 and 1.2 for the five samples (pure CS and four insulin-loaded CS-amide derivatives), and the corresponding correlation coefficient (R²) values are tabulated in

Table 3. Comparison of correlation coefficient (R²) values among the five different pharmacokinetic models for the samples at two different pHs of 7.4 and 1.2.

Sample	Zero Order		First Order		Higuchi Model		Korsmeyer-Peppas Model		Hixson Model
	pH 7.4	pH 1.2	pH 7.4	pH 1.2	pH 7.4	pH 1.2	pH 7.4	pH 1.2	pH 7.4	pH 1.2
Pure CS-Insulin	0.319	0.984	0.364	0.725	0.556	0.958	0.826	0.687	0.348	0.91
CS-g-SA-Insulin	0.889	0.263	0.953	0.326	0.966	0.508	0.835	0.314	0.936	0.274
CS-g-OA-Insulin	0.836	0.215	0.916	0.154	0.967	0.417	0.932	0.259	0.892	0.226
CS-g-LA-Insulin	0.942	0.622	0.972	0.727	0.983	0.869	0.591	0.441	0.976	0.696
CS-g-LLA-Insulin	0.972	0.426	0.969	0.472	0.983	0.667	0.654	0.404	0.986	0.452

. From the comparison of R² values, the insulin release from the pure CS carrier appears to have zero order kinetics at the acidic pH 1.2 (Figure 7a). However, the other samples behave completely different from pure CS, that is, the insulin release from CS-g-SA (Figure 7b), CS-g-OA (Figure 7c), CS-g-LA (Figure 7d), and CS-g-LLA (Figure 7e) trend more towards the Higuchi model at the neutral pH 7.4. In general, the Higuchi model identifies the drug release behavior from matrices, and in our case, the observation of insulin release from amide-derived CS matrices confirms the diffusion process that is based on Fick’s law. Further, the insulin release from the CS-g-LLA carrier at the same pH 7.4 seems to follow the Hixson model as well (R² of 0.986; Figure 7f), and this shows that the release is due to dissolution.

In general, the drug release from a polymer matrix occurs through three different phenomena, which include diffusion-controlled (drug diffusion from a non-degraded matrix), erosion-controlled (drug release due to polymer degradation and erosion), and swelling-controlled (drug release due to polymer swelling). Taking into consideration of these points, the drug release from biodegradable polymers occurs via the cleavage of hydrolytic chains, which allows for the erosion of matrixes and associated diffusion of their loaded drug into the media. However, in the case of non-biodegradable polymers employed as drug-carrier molecules, the drug release occurs through diffusion and matrix swelling. Therefore, resolving those facts to our present scenario, the CS and its cross-linked fatty-acid-derived matrices are considered to be degradable polymers (depending upon the pH of the media). In this way, insulin drug release from these biodegradable matrices at pH 7.4 occurs primarily by diffusion (Higuchi model for all cross-linked fatty-acid-modified CS), and by a combination of both diffusion and dissolution in the case of CS-g-LLA matrix (Higuchi and Hixson models). However, for the pure CS matrix, insulin release mainly follows the zero order kinetic, which indicates that the constant release and drug release rate is independent of the concentration.

The available methods of insulin administration include injectable, nasal, transdermal, and transbuccal. The most proposed one is the oral route, so we incorporated this technique in the present work [25]. It was reported in the literature that a CS/insulin release system can be used for insulin release over a time exceeding 150 h; the release study was conducted in vivo with a hydrogel system using an injectable administration method [26]. Similarly, in another study, a microgel of a CS/insulin release system was tested in vitro and in vivo with a release time of 240 h using an injectable mode of administration [27]. Despite the long periods of insulin release in the aforementioned studies, they used the injectable administration method, which can result in identical problems. Unlike this route, one group developed a CS/insulin system for oral use, but the load capacity of the system has not been revealed [28]. Therefore, considering the view of the oral route and improving the controlled release behavior of insulin drug, the present study successfully tested the release from CS–fatty-acid-based cross-linked matrices under harsh pH conditions and high load capacities, and with a reasonable release time. To the best of the author’s knowledge, this is the first study of its kind where CS–grafted fatty acid matrices with variations in unsaturation level have been studied for insulin release.

4. Conclusions

In conclusion, the study investigates a suitable carrier system for an efficient and controlled delivery of insulin drug from the cross-linked CS-g-LLA amide matrix. Among the tested CS amides with cross-linking, the one with a high amount (3 double bonds) of unsaturation was found to be superior for insulin loading, as well as for releasing carriers, compared to the other amide derivatives of CS with low levels of unsaturation (0–2 double bonds). In addition, compared to the amide derivatives of CS, the insulin loading and release from pure CS were found to be better only under strong acidic environments. The analysis indicated that the formation of a carrier system with fatty acid derivatives with a higher number of unsaturated groups produce some influential changes to the morphological characteristics of CS biopolymer, which allow for a high level of drug loading followed by its efficient release. Further, the studies of pharmacokinetic models confirm that the drug release occurs primarily through the Higuchi model for the amide derivatives, meaning they are diffusion controlled, while zero order in the case of pure CS means that it is independent of all the parameters. Finally, the study provides information about the selection of suitable fatty acids for the development of an effective insulin delivery system, which makes use of CS biopolymer as a base carrier.