Chitosan-Based Coacervate Polymers for Propolis Encapsulation: Release and Cytotoxicity Studies

Chitosan-DNA (CS-DNA) and Chitosan-Pectin (CS-P) hydrogels were formulated as a sustained drug delivery carrier for drug delivery. For this, hydrogels were prepared by emulsion technique: mixing aqueous phase of the CS and DNA or P solution with benzyl alcohol using a high-performance dispersing instrument. Green Propolis (GP) was incorporated by imbibition: hydrogels were placed in GP aqueous solution (70 µg/mL) for 2 h. The specimens were freeze-dried and then characterized using different techniques. In vitro cell viability and morphology were also performed using the MG63 cell line. The presence of P was evidenced by the occurrence of a strong band at 1745 cm−1, also occurring in the blend. DNA and CS-DNA showed a strong band at 1650 cm−1, slightly shifted from the chitosan band. The sorption of GP induced a significant modification of the gel surface morphology and some phase separation occurs between chitosan and DNA. Drug release kinetics in water and in saliva follow a two-step mechanism. Significant biocompatibility revealed that these hydrogels were non-toxic and provided acceptable support for cell survival. Thus, the hydrogel complexation of chitosan with DNA and with Pectin provides favorable micro-environment for cell growth and is a viable alternative drug delivery system for Green Propolis.


Introduction
Hydrogels are known to be crosslinked polymeric 3D networks capable of absorbing and retaining large amounts of water or biological fluids [1], having a microstructure similar to extracellular matrix (ECM) [2] and a biocompatibility with blood and biological tissues [3]. Additionally, hydrogels can be designed with tunable/responsive properties that make them highly sought after for biomedical and pharmaceutical (drug/gene deliver) applications [4,5].
Thus, many efforts have been directed towards hydrogel-based matrices for the development of vectors for the delivery of active substances for different targets. Among different strategies, the use of combinations of natural polymers [6] and biopolymers is mandatory. Following this approach, chitosan is a cationic biopolymer, obtained from the chitin deacetylation of crustaceous exoskeletons [7], showing antimicrobial biodegradable and biocompatible properties. Furthermore, due to the presence on its structure of amine and hydroxyl groups [8,9], the chitosan is highly chemically versatile, being used for several applications in pharmaceutical and medical areas [10], especially for wound healing applications [11], as they are similar to the natural extracellular matrix (ECM). Figure 1a,b shows the FTIR spectra of the CS-P and CS-DNA blend hydrogels, respectively, and the corresponding isolated polymers. The IR spectra of chitosan show two characteristic bands at 1648 cm −1 and 1538 cm −1 assigned to (C=O) stretching mode and N-H bending of amide group. A band at 1378 cm −1 can be attributed to the -CH3 symmetrical deformation and the strong band at 1076 cm −1 characterizes the C-O stretching [38,39]. Furthermore, the spectra show bands at 3417 cm −1 that indicate dimeric O-H stretching and at 1425 cm −1 indicating aromatic C-C stretching [40].

Gel Characterization
To confirm the occurrence of the coacervation process, the presence of pectin was evidenced by the occurrence of a strong band at 1745 cm −1 , also occurring in the blend, ascribed to esterified groups, and at 1624 cm −1 , assigned to free carboxyl groups [41]. The bands occurring at 1105 and 1010 cm −1 , in the finger-print region, are typical of pectin polymers and can be attributed to the C=O stretching [30,42].
In the DNA and CS-DNA gels it is possible to observe a strong band at 1650 cm −1 , slightly shifted from the chitosan band, assigned to the C2=O2 stretching vibration of cytosine [43]; the shoulder found at the blend and at DNA gels, at 1590 cm −1 , can be attributed to a couple of C6=O6, C5-C6, and C4=C5 stretching vibrations of guanine [44]. The small modifications of the FTIR spectra around 1590 cm −1 may suggest a different base-pairing pattern of DNA alone and in the blend [43]. Further bands, in the CS-DNA spectrum, at 526 cm −1 and 890 cm −1 , the latter only observed in the DNA-containing gels, indicate the deoxyribose region [45], and the deoxyribose-phosphate groups [46]. Figure 1a,b shows the FTIR spectra of the CS-P and CS-DNA blend hydrogels, respectively, and the corresponding isolated polymers. The IR spectra of chitosan show two characteristic bands at 1648 cm −1 and 1538 cm −1 assigned to (C=O) stretching mode and N-H bending of amide group. A band at 1378 cm −1 can be attributed to the -CH3 symmetrical deformation and the strong band at 1076 cm −1 characterizes the C-O stretching [38,39]. Furthermore, the spectra show bands at 3417 cm −1 that indicate dimeric O-H stretching and at 1425 cm −1 indicating aromatic C-C stretching [40].  Excluding the T max obtained at temperatures below 150 • C, caused by the evaporation of residual water, two main degradation steps were found at 234 • C and 292 • C for pectin [47,48] and chitosan [49], respectively. It is interesting that for the CS-P blend, two degradation steps can be found at 227 • C and a 548 • C; the latter can be justified by the degradation of chitosan. The former maximum degradation temperature, corresponding to ca. 63% weight loss, occurs at temperatures below the main transition steps for single pectin and chitosan. This can be justified by a decrease in the thermal stability of the blend as a consequence of ionic interactions between pectin and chitosan, leading to a plasticizer-like effect.
The gels have also been characterized by thermogravimetric analysis (Figure 2a,b). In order to assess the main degradation temperatures, the maximum degradation rate was computed (dTG), and the corresponding temperature, T max , obtained.
The analysis shows that the first degradation step of DNA occurs at 238 • C [50], followed by a second degradation step at 295 • C, suggesting that the presence of different secondary structures might be present. However, the most remarkable difference between this blend, and the CS-P one, is that the main degradation step occurs at 268 • C, suggesting a higher degree of phase separation.
From the analysis of the surface morphology of CS-P and CS-DNA gels ( Figure 3), it can be observed that the surface of the latter shows a more heterogeneous structure, characterized by well-defined aggregates and fiber-like structures. On the other hand, for CS-P, the surface morphology is a more featureless, non-porous structure, indicating some irregularities that might be related to some phase separation. This morphology can increase stability by providing higher resistance [51]. However, after the GP sorption, a significant change in the surface morphology of CS-DNA is observed, i.e., the surface becomes highly porous. This has been reported for other DNA-containing composites, such as PVA-DNA [43], and suggests that the GP has a significant effect or even interaction with the polymers. This might also justify the lowest amounts of GP release by CS-DNA (see the discussion below), when compared with the CS-P, despite its higher porosity. However, the surface of CS-P does not undergo relevant modification upon GP loading, indicating that the pectin is playing a major role [52].
for single pectin and chitosan. This can be justified by a decrease in the thermal stability of the blend as a consequence of ionic interactions between pectin and chitosan, leading to a plasticizer-like effect.
The gels have also been characterized by thermogravimetric analysis (Figure 2a,b). In order to assess the main degradation temperatures, the maximum degradation rate was computed (dTG), and the corresponding temperature, Tmax, obtained.  observed, i.e., the surface becomes highly porous. This has been reported for other DNA-containing composites, such as PVA-DNA [43], and suggests that the GP has a significant effect or even interaction with the polymers. This might also justify the lowest amounts of GP release by CS-DNA (see the discussion below), when compared with the CS-P, despite its higher porosity. However, the surface of CS-P does not undergo relevant modification upon GP loading, indicating that the pectin is playing a major role [52].

Loading and Drug Release
A wide variety of polymer-based materials have been developed for drug delivery, such as biodegradable chitosan hydrogels [53]. Table 1 shows the encapsulation efficiency of GP into CS-P and CS-DNA blend gels.

Loading and Drug Release
A wide variety of polymer-based materials have been developed for drug delivery, such as biodegradable chitosan hydrogels [53]. Table 1 shows the encapsulation efficiency of GP into CS-P and CS-DNA blend gels. It can be seen that the EE in the CS-DNA is significantly lower (ca. 3.5-fold lower) than those obtained for CS-P. This might be related to the occurrence of higher phase separation in the CS-DNA gel and also the lower swelling degree of this gel (Q = 700%) when compared to that of CS-P (Q = 1109%). However, it is surprising that the cumulative release, in percentage, is higher for the CS-DNA gel. This might be due to the change in the structure of the pectin-DNA gel, which becomes a more porous surface structure. In any case, the absolute value of the maximum released amount of GP is always higher for CS-P. Figure 4 shows the release kinetics of GP from CS-P gels to water (diamond) and simulated saliva (spheres), at 25 and 37 • C (black and white data points, respectively; Figure 4a,b). The solid lines were obtained by fitting Equation (1) to the experimental data. Figure 4c,d shows the release kinetics of GP from CS-DNA gels to water (triangle) and simulated saliva (square), at 25 and 37 • C (black and white data points, respectively). Solid lines were obtained by fitting Equation (1) to the experimental data.  To obtain better insight into the release mechanism, several heuristic models were tested (e.g., Weibull, Logistic, Korsmeyer-Peppas and Peppas-Sahlin [21,55,56] without success-results not shown. However, assuming that the release is characterized by a biphasic release that may occur simultaneously, as can be inferred from the continuous increase of the GP as a function of time (i.e., no two plateaus of cumulative release concentration can be observed as a function of time), the following equation was used [57]: where C0, Ct and Ceq are the initial, cumulative and equilibrium release concentrations, respectively, of GP, logt1 and logt2 correspond to the first and second EC50 (i.e., the concentrations of As a first approach, we can observe that in all cases, the release kinetics follows a two-step mechanism; however, it is also clear that such a mechanism occurs concomitantly when no plateau is observed just before the beginning of the second significant release-this will be discussed later. Two other points are worth noticing: the cumulative release increase with increasing temperature (the exception occurs for the release from CS-P to water), which is more significant in the case of the CS-P gel; and the boost release that can be observed during the first 10 min of release at 25 • C. The latter process might be justified by a surface phenomenon (either surface erosion or preferential adsorption) [54]; however, that process is somewhat hindered by increasing the temperature and, consequently, the origin of mass transport ceases to be relevant.
To obtain better insight into the release mechanism, several heuristic models were tested (e.g., Weibull, Logistic, Korsmeyer-Peppas and Peppas-Sahlin [21,55,56] without success-results not shown. However, assuming that the release is characterized by a biphasic release that may occur simultaneously, as can be inferred from the continuous increase of the GP as a function of time (i.e., no two plateaus of cumulative release concentration can be observed as a function of time), the following equation was used [57]: where C 0 , C t and C eq are the initial, cumulative and equilibrium release concentrations, respectively, of GP, logt1 and logt 2 correspond to the first and second EC 50 (i.e., the concentrations of the drug that give half-life response), and p is the contribution of each phase for the whole release profile. All the fitting parameters, summarized in Table 2, were estimated on the basis of a least-squares method, with a confidence degree of 95%, by using Origin 8.5 software, while keeping C 0 as a constant equal to zero. From the analysis of Figure 4a,b and C eq summarized in Table 2, it can also be concluded that the amount of GP released is significantly higher in synthetic saliva than in water. The amount of GP released reaches a value between 14% and 36% of the encapsulated amount of GP. This may be attributed to the effect of ionic strength on these gels. Once both gels are ampholytic gels, an increase in the ionic strength may lead to a swelling effect, which may stabilize both polymers [58,59], thus contributing to the enhancement of GP release. This is also followed by a significant increase in the EC 50 for the first step release (logt 1 ) compared to those observed for the water system. Finally, concerning the contribution of both steps for the whole release process, the first step mechanism contributes ca. 30% for the total release of GP for all systems, suggesting that the main release process is probably affected by some polymeric relaxation processes, further facilitating the mobility and release of the GP [60-62].

Biological Analysis
The results of cell viability MTT assay and total protein production are shown in Figure 5. With respect to total protein (Figure 5b), there was no statistical difference between the control group and CS-DNA (p > 0.001), but the values from the CS-P experimental group were lower than those for the control group, with statistical significance (p = 0.0002).
The cell viability was obtained by MTT assay, and measurements of cell viability of experimental groups were compared to the control group (100%) (Figure 5a). The CS-P and CS-DNA were positive influences in the cell culture and improved the number of viable cells. However, the CS-P experimental group exhibited a statistically significant difference with respect to the control group (p < 0.0001), which presented lower cell viability. This result is in accordance with previous studies, which observed an increase of viable cells in different linage cells when chitosan was used as a biomaterial base [63][64][65]. The rates of cell viability exhibited by both groups make them suitable for biomedical use according to the International Organization for Standardization (ISO) and the national standards for Biological Evaluation of Medical Devices, ISO 10993-5 [66]. From the analysis of Figure 4a,b and Ceq summarized in Table 2, it can also be concluded that the amount of GP released is significantly higher in synthetic saliva than in water. The amount of GP released reaches a value between 14% and 36% of the encapsulated amount of GP. This may be attributed to the effect of ionic strength on these gels. Once both gels are ampholytic gels, an increase in the ionic strength may lead to a swelling effect, which may stabilize both polymers [58,59], thus contributing to the enhancement of GP release. This is also followed by a significant increase in the EC50 for the first step release (logt1) compared to those observed for the water system. Finally, concerning the contribution of both steps for the whole release process, the first step mechanism contributes ca. 30% for the total release of GP for all systems, suggesting that the main release process is probably affected by some polymeric relaxation processes, further facilitating the mobility and release of the GP [60-62].

Biological Analysis
The results of cell viability MTT assay and total protein production are shown in Figure 5.   Demonstrating the labeled cellular targets of the hydrogels, the images reveal cell-to-cell (Figure 6a,b,e,f), cell-CS-DNA (Figure 6c,d) and cell-CS-P (Figure 6g,h) contact. These fluorescence microscopic images illustrate the cell distribution on the hydrogel surfaces to be dependent on their migration and proliferation, as distinct from the homogeneous distribution that occurs in process in which the cells are encapsulated inside the gel [67]. experimental group exhibited a statistically significant difference with respect to the control group (p < 0.0001), which presented lower cell viability. This result is in accordance with previous studies, which observed an increase of viable cells in different linage cells when chitosan was used as a biomaterial base [63][64][65]. The rates of cell viability exhibited by both groups make them suitable for biomedical use according to the International Organization for Standardization (ISO) and the national standards for Biological Evaluation of Medical Devices, ISO 10993-5 [66]. Figure 6 shows fluorescence microscopic visualization of cells in contact with hydrogel.

Materials
Chitosan with an acetylation degree of 15 mol% (Mw: 87 × 103 g mol −1 ) was purchased from Golden-Shell Biochemical (Yuhuan, China). Pectin from citrus peel and double-stranded (ds) DNA (sodium salt from salmon testes), with a polymerization average degree of ca. 2000 base pairs, were purchased from Sigma. Benzyl alcohol (99%) was obtained from Merck KGaA (Darmstadt, Germany).
Aqueous extract of Green Propolis was obtained from Apis Flora (Ribeirão Preto, SP, Brazil). Millipore-Q water was used to prepare the solutions. Simulated saliva was made by Indiana University modified method [68], prepared according to Fusayama et al. (1963) [69].

Hydrogel Preparation
Both cocervate hydrogels (CS-DNA and CS-P) were prepared using the emulsion technique (oil-in-water). The preparation of the former gel was done using an acetate buffer (pH 6). This buffer was chosen to balance the cationic density of chitosan and the chemical instability of DNA [70].
Thus, an aqueous chitosan solution (1% w/v) was initially prepared in acetate buffer and then filtered through a paper filter to remove insoluble substances [30]. DNA and pectin aqueous solutions were prepared in phosphate buffer at pH 6 and pH 9.2, respectively, under stirring for 12 h, at room temperature. The emulsions were obtained by mixing aqueous phase of the Ch solution (1 mL) and DNA solution (1 mL) (1:1 (v/v) ratio of Ch:DNA) or Pectin solution (1 mL) (1:1 (v/v) ratio of Ch:Pectin) with benzyl alcohol (5 mL) using a Ultra-Turrax at 31-34000 rpm min-1 for 5 min [71].

Hydrogel Characterization
The synthesized hydrogels were freeze-dried prior to their characterization using various techniques. Attenuated reflection infrared spectroscopy (ATR-FTIR) was carried out in a Cary 630 FTIR Spectrometer (Agilent, Santa Clara, CA, USA), with wavenumber ranging from 650 cm −1 to 4000 cm −1 .
The surface morphologies of hydrogels were analyzed using scanning electron microscopy (SEM, JEOL USA, Inc., Peabody, MA, USA) with a JEOL model 5310 scanning microscope (JEOL USA, Inc., Peabody, MA, USA) operating under low vacuum at 15 kV. The membranes were submitted to a fast cryogenic treatment by diving gel samples into liquid nitrogen for 10 s, and then they were left overnight in a freeze dryer (Free Zone 4.5-Labconco) before being coated with a gold film.
The effect of water and simulated fluids on hydrogels before and after the loading of Green Propolis was assessed by swelling degree (Q). Different samples of the same membrane were cut, weighed, and immersed in water or simulated solutions and left to reach swelling equilibrium (2 days). After this time, membranes were removed from the solution, any drops of solution were wiped off, and then the weight of the membrane was measured using an analytical balance ADA 120LE (Sartorius AG, Göttingen, Germany). Each experiment was repeated twice. The swelling degree of the gel (Q) was calculated by using the following equation: where m w and m x are the masses of water (or solution) in the swollen gel and in the xerogel, respectively. The mass of the xerogel samples was obtained using the weight of the sample after synthesis and taking into account the solid content of the gel [43].

Loading and Release of Green Propolis
Green Propolis was incorporated into the coacervate gels by imbibition. A 70 µg/mL GP aqueous solution was prepared; after that, hydrogel samples (of ca. 20 mg) were placed in that volume of GP solution and left there for 2 h, at 25 • C, using a thermostatic bath.
The encapsulation (EE) efficiency was calculated by using, respectively, the following equations: To determine the GP release kinetics, the hydrogel samples, after GP encapsulation, were placed in vessels containing 20 mL of water, and kept in an incubator (Labwit, ZWY-100H, Burwood East, VIC, USA), at 37 • C and 150 rpm. At predetermined time intervals (20 min), aliquots of the supernatant were taken from the vessel and stored for GP quantification. Meanwhile, that volume was replaced with water. The process was repeated until release equilibrium was attained. The GP released into the supernatant was quantified spectrophotometrically by measuring the absorbance in a range of 200 to 800 nm using a double-beam Shimadzu UV-2100 spectrometer (SHIMADZU EUROPA, Duisburg, Germany).

Biological Analysis
In this present study, an established lineage of osteoblast-like cells (MG63) acquired from Rio de Janeiro Cell Bank (Rio de Janeiro, RJ, Brazil) was used. Cells were cultured in Modified Eagle's medium (Life Technologies, NY, USA) containing 10% fetal bovine serum (Life Technologies, NY, USA) and penicillin (100 U/mL) at 37 • C in a humidified atmosphere of 5% CO 2 . The culture medium was changed every two days until the moment test. All tests were performed according to ISO 10993 guidelines, with three independent experiments.
Initially, five samples for each material were placed into 24-well plates (TPP, Curitiba, Brazil), with 2 × 104 cells subsequently being placed in each well. After 7 days for culture, an MTT colorimetric assay was performed with incubation in 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma-Aldrich, St Louis, MO, USA); then, the cells were lysed with a propanol acid solution (Sigma-Aldrich, St Louis, MO, USA) for colorimetric measurement in a spectrophotometer (EL 808 BioTek Instruments, Winooski, VT, USA) at 570 nm, as previously described [72]. The values were expressed as percentage of viable cells compared to the control group (only cells), which was considered to be 100%.
Proteins were extracted with 0.1% sodium dodecyl sulfate (Sigma-Aldrich St Louis MO, USA), followed by addition of Lowry solution (Sigma-Aldrich, St Louis, MO, USA). The extract was diluted in Folin-Ciocalteu reagent (Sigma-Aldrich, St Louis, MO, USA) and the absorbance was assessed in a spectrophotometer (8582 Micronal, São Paulo, Brazil) at 680 nm [72]. The total protein content was calculated from a bovine serum albumin standard and expressed in µg/mL.
Finally, for cell morphology of the actin cytoskeleton and cell nucleus adhered, the cultures grown on the samples for 3 days and were fixed for 15 min using 4% paraformaldehyde (Synth, São Paulo, Brasil) in 0.1M sodium phosphate buffer (PBS) (pH 7.2), at room temperature. Briefly, they were permeabilized with 0.5% Triton X-100 (Sigma Aldrich, St Louis, MO, USA) in PB for 15 min, followed by 1% ovalbumin for 30 min, and Alexa Fluor 488 Phalloidin (Invitrogen Eugene, Oregon, USA) was subsequently added for 20 min, followed by two washes with PBS [73]. The analyses were performed on the well cells that were in contact with the samples under the fluorescence microscope (Microscope Carl Zeiss Microlimaging GmbH-Axiovert 40C, Germany). At the end of the procedure, Fluoroshield with DAPI (Sigma-Aldrich, St Louis, USA) was added. The digital images were visualized using the Axio Vision 7.0 software.

Conclusions
This study reports a necessary bridge between chemistry and biology, since the control of properties is imperative for in vivo applications of new biomaterials. To achieve this, biocompatible coacervate hydrogels based on chitosan, with either salmon DNA and pectin, were synthesized by the emulsion technique for drug delivery of the herbal medicine Green Propolis. In addition to a chemical and morphological characterization, the release transport of GP that had been previously encapsulated into gels in different media and at different temperatures were assessed along with the gel's cytotoxicity.
The hydrogel release kinetics, analyzed at different temperatures (37 • C and 25 • C) and with different vehicles (water and saliva), exhibited a two-step mechanism, a cumulative release that increased proportionally to the temperature, significantly for the CS-P, with a boost release at 25 • C. Moreover, the amount of GP released was higher in saliva than in water.
Regarding cell viability, both hydrogels positively influenced the cell culture and improved the number of viable cells. CS-DNA showed classical growth of total protein assay, statistically similar to the control group and different from CS-P.
Considering these factors, this study reports the successful fabrication of hydrogels composed of chitosan-DNA and chitosan-pectin, potential alternatives to drug delivery systems, and preliminary evidence suggesting that both proposed matrices are favorable micro-environments for potential use as materials in the biomedical field.