Superabsorbent biodegradable CMC membranes loaded with propolis: Peppas-Sahlin kinetics release

Abstract Propolis is a resinous product collected by honeybees with a complex chemical composition. Sodium carboxymethylcellulose is a polymer commonly used in wound care. The goal of the present work was to produce and characterize NaCMC membranes loaded with extract of Brazilian brown propolis (CMC-P). Flavonoids and phenolic acids were identified in the propolis extracts, where the main identified substance was kaempferide. The brown propolis extracted was active against S. aureus. The low swelling capacity and high gel fraction of CMC-P would be the consequence of propolis (responsible for a hydrophobic barrier) filling the pores of the membrane. Propolis could be anchoring the NaCMC chains (as observed by FTIR) due to interaction between components, which is corroborated by the CMC-P sample degrading less than the CMC sample (>400ºC). There was non-linear diffusion release kinetics for most phenolic substances of the propolis extract. The CMC-P sample presents potential as a dressing material.


Introduction
Wound proper care should reach the previous steps without delay.However, the World Health Organization highlights the importance of wound proper care to avoid infection, since up to 30% of patients in intensive care develop an infection. [3][4]Wound infection can be considered a major cause of healing delay and high costs related to it. [5]Regarding skin wound infection microorganisms, P. aeruginosa (gram-negative) and S. aureus (gram-positive) are common bacteria that colonize wounds. [1]Infection control might be difficult since to avoid bacterial resistance, the rational usage of systemic antibiotics is advised. [6]Infection kinetics is also a variable process, but usually, Gram-positive bacteria are the first ones to colonize the wound site (1 st week of infection), followed by gram-negative bacterial colonization.Regarding P. aeruginosa and S. aureus resistance, they usually are resistant to several antibiotics, e.g., methicillin, [7] carbapenems, [8] cephalosporins (3 rd generation antibiotics) [9] .
Bee products like honey are known to inhibit both, grampositive and gram-negative resistant mechanisms, due to their broad spectrum of activity. [10]In addition, propolis, at proper amounts, can be bacteriostatic and even bactericide on wound infections without inducing bacterial resistance. [11]ropolis is a beehive product containing mainly beeswax and resins obtained from plants.Apis Mellifera bees produce propolis to seal the hive, protecting it from insects and pathogenic microorganisms. Propolis presents antibacterial, antioxidant, antifungal, anti-inflammatory, and wound-healing properties. [14] The color of propolis varies from dark green to reddish, and its composition depends on the plant source of the resin.The most studied Brazilian propolis is the green type originating from Baccharis dracunculifolia and produced mainly in the southeast of Brazil. [12]Other types of Brazilian propolis are known, such as brown and yellow ones, but with undetermined plant origin.Although its geographical origin is unknown, Brazilian brown propolis is rich in terpenes, [17] while Brazilian yellow propolis can be considered rich in triterpenoids. [18]opolis' effects on wound healing are according to its composition.Nonetheless, flavonoids' bactericide mechanism includes damage to bacteria's cytoplasmic membrane and inhibition of nucleic acid synthesis. [19]egarding propolis application in wound care, green and red propolis have been successfully applied in Wistar rats' wounds.Although red propolis presented high amounts of flavonoids, green propolis led to high reepithelization, [20] also controlling inflammatory response. [21]Red propolis was tested in male Wistar rats' wounds.It increased the wounds' contraction rate, as well as stimulated healing factors, and increased collagenase activity. [22]Brown propolis was compared to green propolis regarding their action on oxidative stress and inflammation.Brown propolis and green propolis present different metabolite profiles and mechanisms of action, but brown propolis was more active than a green one. [23]Brown propolis was tested against the biofilm formation of S. aureus.Propolis' prenylated phenylpropanoic acids were antibiofilm (S. aureus' colonies spread) probably due to artepillin C, drupanin, and baccharin metabolites. [24]Brown propolis was incorporated in alginate membranes to be used as a food covering when the membrane was active against gram-positive bacteria. [25]dium carboxymethylcellulose (CMC) is one of the main polymers derived from cellulose.It is soluble in water, and its viscosity and adsorption capacity can be modulated by varying pH, concentration, and temperature.It is a hydrocolloid, forming a gel or a viscous dispersion in water. [26][27]Sodium carboxymethylcellulose has recently been used in dressings for the treatment of wounds and burns.Its membranes improve the healing process, and they can be used pure or in combination with other polymers.Due to its hygroscopic characteristic, carboxymethylcellulose promotes an autolytic debridement of wounds; facilitates cellular rehydration; and has bacteriostatic action. CMC hydrocolloids allow the incorporation and controlled release of different drugs or natural products.CMC-tamarind gum hydrocolloids were loaded with moxifloxacin hydrochloride, where the equimolar gel delivered moxifloxacin hydrochloride properly [31] .Red propolis extract from Alagoas/Brazil was added to CMC hydrogels, and they were effective against microbes' penetration towards the wound site, being considered promising materials for dressings. [32]The goal of the present work is to develop and characterize CMC membranes loaded with Brazilian southeast brown propolis for wound care.

Propolis analysis
Propolis extraction was performed by four methods: dynamic maceration at room temperature (DM) -50ml of ethanol PA was used as solvent for 48h to extract 2g of propolis; dynamic maceration at 50ºC (DMT) -50ml of ethanol PA was used, at 50ºC for 48h, to extract 2g of propolis; by ultrasound bath (US) -50ml of ethanol PA in ultrasound bath for 2h at room temperature; and by immersion of ultrasound probe (USI) -50ml of ethanol PA under ultrasound probe for 30min at room temperature.The samples were then characterized following their active compound amounts.

Propolis antioxidant activity (DPPH, FRAP, and ABTS)
The percentage of antiradical activity was calculated through the decolorization of the DPPH• radical (Sigma-Aldrich), according to Equation 1. [34] To determine the antioxidant activity (%AA) of each propolis sample, the absorbance of the solution of methanol + DPPH (Abs control ) was considered the negative control, and methanol was used as a reference (Abs REF ).This analysis was performed in triplicate.To calculate the amount of antioxidant (hydrogen donator) necessary to diminish the initial concentration of DPPH by 50% (CE 50 ), a stock methanolic solution of propolis was prepared (concentration of 1000 μg/mL).Then, 0.29μL of the DPPH solution was added to ELISA plate wells as well as the prepared solutions.After 30 minutes of incubation in the dark, the absorbance was measured at 520nm.Analysis was performed in triplicate.

CLAE-DAD chromatography
Chromatographic analysis was performed on a C-18 reversed-phase analytical column (Betasil, Thermo, 5μm particles size), at 30 °C.The mobile phase used was water with 1% acetic acid (solvent A) and methanol (solvent B), a constant flow of 1.0 mL/min, and a volume of the injected sample of 20μL.The concentration gradient was performed from 35% of solvent B for 2min, followed by (35-80)% at 20min, (80-92)% at 25min, returning then to 35%B for 2min.The chromatograms were recorded at 280 and 340 nm since most of the phenolic acids and flavonoids in propolis are excited near these wavelengths.Data acquisition used the LCSolution (Shimadzu) software.

Antimicrobial analysis
The antimicrobial activity of propolis extracts against S. aureus was determined according to the agar diffusion method with modifications. [36]The S. aureus strain (ATCC 25923) was incubated at 36ºC for 24h.A suspension of cells was prepared in 3mL of peptone saline solution to reach 2x10 8 cfu/mL (turbidity equivalent to Mc Farland scale number 5).This suspension was diluted 100 times and 0.1 mL was inoculated on Mueller Hinton agar plates.Wells (sterile drill, diameter of 0.75 cm) were drilled and, in each well, 50 µL of propolis extract (10 µg/mL) was inoculated.Ampicillin (10 µg/mL, positive control) was used, as well as a solution of 95% ethanol (negative control).The incubation occurred at 36°C for 20-24 hours, after which the inhibition diameter of each well was measured.All tests were performed in triplicates.The inhibition (I) was calculated according to Equation 2.

( )
Sample s inhibition halo Negative control s halo I Positive control s halo

CMC gels
The hydrogels were prepared according to the casting method [37] .3g of sodium carboxymethyl cellulose -CMC (Sigma Aldrich) was diluted in 100 mL of deionized water under mechanical stirring (Fisatom brand, model 710) for 90 min at room temperature.The solution was then divided into 25 mL portions and, under magnetic stirring (Fisatom brand, model 752), 0.8g of citric acid (VETEC) was added [38] [39] .The samples were dried in an oven (Nova instruments, model NI1512) for 24 h at 50 ºC.The CMC membranes obtained were immersed in deionized water for excess citric acid removal (10 mL of deionized water for 24 h in an oven at 50 ºC per sample).After this, the CMC membranes were subjected to the swelling process in propolis extracts (20 w/v%, resulting in 2.4 g of propolis per membrane) for 24 h.

Physical properties
The samples' physic-chemical analysis was conducted by FTIR.The analysis was performed on Bruker equipment (Vertex 70), in the range of 400 cm -1 -4000 cm -1 , 16 scans per sample.The microstructural analysis of the samples was performed by X-ray diffraction (XRD) using the equipment Brucker-AXS D8 Advance Eco diffractometer (CETEM -UFRJ), with CuKα radiation at 40 kV and 25 mA, angular diffraction range of 2θ = 5°-70°, a step of 0.02°, and step time of 2 seconds.The XRD plotted curves were then smoothed (Method Savitzky-Golay, 400 points per window).The thermogravimetric analysis used the equipment TGA Q500 (TA Instruments Co.), Catalysis Lab-UFRRJ.The samples (~10 mg) were loaded in an open platinum crucible, where an empty crucible was used as a reference, continuous flow of N 2 (30 mL.min -1 ), a heating rate of 10 °C / min, between 25° C and 400° C [40] .For the mechanical tests (modified ASTM D882 -02 standard), rectangular samples of (40x20x1) mm 3 in triplicate were evaluated (EMIC DL 10000 equipment, load cell Trd 21 / 50 kgf), strain rate of 3 mm/min until failure, at room temperature.

In-vitro properties
The swelling test was performed in triplicate, where the hydrogels were immersed in 25 mL of saline solution [41] .The sample weight was evaluated at predetermined time intervals (0.5h, 1h, 2h, 3h, 4h, 24h, and 48h of immersion).The samples' swelling degree (SG) was calculated according to ( ) % 100( / ) , where W t is the weight of the samples at each time interval and W i , is the dry samples' initial weight.The samples' gel fraction (GF) and weight loss (WL) were calculated according to ( ) ( ) , respectively, where W f is the final dry weight of the samples (after 96h of swelling).
The in vitro release study of phenolic compounds was carried out according to the shake-flask methodology [42] .Carboxymethylcellulose membranes impregnated with propolis extract were immersed in 100mL of phosphate buffer pH 6 with 1% Sodium Sulfate Luaril (Sigma Aldrich).They were stirred under 50 rpm for 96 h, at 32ºC.The aliquots of the samples were collected at the following times: 0.5; 1,0;2,0; 3,0; 4,0; 5,0; 24; 48; 72; 96h for analysis of the release kinetic profile by HPLC-DAD.Before analysis, the aliquots were filtered (0.45 μm membrane) and analyzed by HPLC.
The data acquisition was done through the LCSolution software (Shimadzu).The analyses were performed in a reverse phase analytical column C-18 (Betasil, Thermo), maintained at 40 ºC.The mobile phase used was ultrapure water with 1% acetic acid (solvent A) and methanol (100, solvent B), with a constant flow of 1.0 mL min/4.6 mm, and the volume of the sample injected was 20μL.The propolis aliquots were solubilized in spectroscopic grade methanol, with a concentration of 1000 μg/mL, and the solution was filtered (0.45 μm, PVDF, Millipore).The substances identification was based on the comparison of retention times.The release kinetics of propolis substances was performed through non-linear regression analysis, applying two models: Korsmeyer-Peppas and Peppas-Sahlin models.The model with the highest R 2 value was considered the best fit.

Propolis
Brown propolis from Southeast Brazil was evaluated and its flavonoids varied from 37 mg quercetin/g propolis extract (MDT) up to 58 mg quercetin/g propolis extract (US), Figure 1.The minimum concentration of flavonoids required by Brazilian regulation is 0.25% (w/w). [43]All propolis extracts presented higher amounts of flavonoids than the minimum required.Nonetheless, the current propolis presents low amounts of flavonoids, since brown propolis from northeast Brazil presented ~14% flavonoids. [25]he samples also presented phenolic acids (from 45 mg gallic acid / g propolis extract (DM) up to 100 mg gallic acid / g propolis extract (DMT), 4.5-10% phenolic acids), but this amount can be considered low. [25][44]Maceration can be considered the best process, Figure 1, although the ultrasound energy usually results in high content of phenolic acids and flavonoids in the propolis extract. [45]ll samples presented antioxidant activity/scavenging activity.Regarding brown propolis, [46] low amounts of propolis extract were required to scavenge 50% of DPPH (from DMT (3.7 μg/mL) to USI (4.1 μg/mL)).The ABTS scavenging activity was in the expected range, [47] from 91 mmol trolox/100mg of propolis (DM) up to 103 mmol trolox/100mg of propolis (USI).DMT and USI techniques presented high FRAP results (264 mmol Fe(II)/100 mg propolis extract (DM) -372 mmol Fe(II)/100 mg propolis extract (USI)), although these values can be considered low for propolis extracts. [46]The ultrasound extraction technique presented high values of antioxidant activity according to DPPH, FRAP, and ABTS methods. [48]egarding the identified phenolic compounds, there were chlorogenic, caffeic, ferulic, para-coumaric and rosmarinic acids, pinobanksin, kaempferol and kaempferide.The phenols amount in the propolis ethanolic extracts differ considerably, e.g., propolis from Brazilian South's region may present low amounts of artepilin C and coumaric acid. [18]Chromatographic analysis revealed the main active compounds in the propolis extract, Figure 1.Kaempferide was the main flavonoid in the propolis extracts, regardless of the type of extraction. [49]here were lower amounts of rosmarinic acid [50] , p-coumaric acid [51][52] , ferulic acid [53] , and pinobanksin [54] , mostly substances presenting antioxidant activity.Propolis extract also presented low amounts of chlorogenic acid, kaempferol, and caffeic acid. [55]ropolis extract obtained through various techniques did not present significant differences regarding the active compounds.Then, macerated propolis extract was the one evaluated by antimicrobial activity.Propolis was placed in agar discs' wells, where S. aureus was incubated.The samples presented measurable halos, inhibiting approximately (40.97 ± 2.41) % of S. aureus microorganisms.The antimicrobial activity of propolis extracts is often related to their phenolic content. [56]Phenolic substances are capable of interfering with the structure and properties of bacterial membranes, increasing their susceptibility to proton permeation, and resulting in microorganisms' death. [57]The propolis extract activity against S. aureus might be related to the high amount of kaempferide, an antimicrobial substance related to a skin infection. [58]

CMC gels
Since brown propolis obtained through different techniques were similar regarding the active compounds, the direct macerated (DM) extract was used to incorporate in CMC membranes.Swelling capacity is a fundamental characteristic of hydrogel dressings, since they can absorb wound exudate. [59]Hydration content and swelling are relevant properties of dressings.Healing is improved by a moisturized environment.Wound dressings with high swelling capacity would be the best ones, they would absorb exudates and diminish the occurrence of infection. [60]MC membranes presented high water absorption capacity compared to CMC-propolis (CMC-P) samples, Figure 2.Both samples reached the equilibrium of swelling degree (ESD) in 24h, CMC samples reached an ESD of ~449% while CMC-P samples reached an ESD of ~168%.[61] , where CMC gels could be considered superabsorbent.[62] The low swelling capacity of CMC-P would be the consequence of propolis filling the pores of the membrane.Since the ethanolic extract of propolis is quite resinous, it may be responsible for a hydrophobic barrier formation.This barrier difficult the absorption of moisture by the pores of the polymer, leading to a low swelling degree and low weight loss.[63] The CMC-P samples presented a high gel fraction, where the presence of propolis extracts difficult the polymer's chains mobility, interfering with the entanglement of amorphous chains.Due to the hydrophobic barrier of propolis, fewer chains could be leached out by the water entrance, and the CMC-P samples' weight loss was low, as well as its biodegradability rate in water.[64] Propolis extract, rich in phenolic substances, may be forming bonds with the polymer chains and thus contribute to the physical crosslinking of the gels, hindering the expansion of CMC chains.A low swelling degree allows a slow release of the active agents.[65] The CMC sample, as well as the CMC-P sample, presented these materials' characteristic FTIR bands and vibration modes, in Table 1 and Figure 3.The CMC-P sample presented mainly bands related to both phases, Table 1.CMC and propolis bands and their vibrational modes; CMC-P samples bands.
Regarding the XRD analysis, the CMC and CMC-P diffractograms were smoothed, then Gauss curves were fitted.Three crystalline peaks (2θ ~11º, ~21º, and ~35º) were identified, in Figure 3, related to CMC diffraction plans (110), (200), and (004) respectively. [82][83]The data (Half-width of the peak -FWHM, location of the peak, 2θ (rad)) related to the main peak (2θ ~21º) were the basis to calculate, by Scherrer equation, the crystallite size of CMC. [84]The CMC sample presented a crystallite size of 9.2 Å while the CMC-P sample crystallite size was 7.4 Å, which can be considered low-size crystallites. [85]Propolis seems to interfere with the bonds between CMC chains, leading to low crystallite size.Probably it breaks hydrogen bonds between CMC molecules; or the presence of propolis physically interferes with the CMC chains' mobility, diminishing the possibility of contact between chains. [86]egarding thermal properties, propolis increased the stability of the samples, Figure 4.The CMC hydrogel thermal degradation analysis began at 25º C, and the slight weight loss between 25 ºC-200 ºC would be related to volatile (H 2 O, etc.) substances.The major loss is at 277 ºC, due to the degradation of the CMC chains (bonds' cleavage related to functional groups and loss of weak groups of the main chains). [87]The degradation profile of both samples, CMC e CMC-P, are similar, but their behavior at high temperatures (above 400 ºC) differs.The CMC-P sample degraded less than the CMC sample, probably due to propolis interaction with the CMC chains (which could be compared to the citric acid effect on CMC), where connection by hydrogen bonds would increase this sample's thermal stability. [88]he CMC sample's final residue was 1% while the CMC-P sample was 23%.The CMC sample weight loss in the first step was ~18% and in the second step, 49%.The CMC-P samples presented 14% and 37% of weight loss in the mentioned steps, where these two first degradation steps would be responsible for the samples' high degradation. [89][90]he increased stability of CMC-P was also shown by the high GF values and by the new FTIR bands, indicating an interaction between components of the CMC-P sample.The CMC sample presented many degradation steps above 400 ºC, which might be related to the products of the 2nd stage degradation step, where the last steps would lead to gases (CO, CO 2, etc.) evolution and carbonaceous residue. [91]he tensile tests of samples (triplicates for each composition) were performed until failure.CMC-P samples presented high strength compared to CMC samples, Figure 4.The samples' Young Modulus (E), elongation at break (e), and Failure strength (σ f ) were evaluated by ANOVA-1 way analysis (factor: composition; levels: CMC and CMC-P), with a confidence level of 95%.It was observed that the CMC sample presented Young Modulus (E) significantly lower than CMC-P (p-value = 0.01732).In addition, according to the ANOVA analysis, CMC-P showed failure strength significantly higher than CMC, with p-value = 0.04781.These results are in agreement with the finds reported in the FTIR and TGA analysis, where the propolis connection to the CMC  chains could anchor the CMC molecules, diminishing the chains' mobility (and elongation at break), and increasing E and σ f .Nevertheless, the samples' failure strength (CMC and CMC-P) can be considered low, as well as the young modulus. [92][93]The CMC-P samples presented elongation at break significantly lower than CMC samples (p = 0.02858).The elongation at the break would indirectly represent the samples' crosslinking, indicating effective interaction. [94]ound dressings have been designed to carry and to release drugs / antimicrobial agents locally.This property would be adequate for chronic lesions treatment, for wounds presenting prolonged inflammation step, as well as for delayed wound closure. [95]Mapping the drug release of the matrix (toward a controlled profile) would guarantee that the drug dose is within the therapeutical level.There are sustainable drug release systems (commercially available), e.g., Insulin Pen e Synchromate B., none of them for wound healing, though. [96]To highlight the importance of kinetics release knowledge regarding wound dressings, drug delivery can be the sine qua non condition to achieve robust delivery's steps and controlled amount of drug locally released.Local antibiotic sustainable release (within the dressing), compared to systemic antibiotic administration, led to efficient bactericide effect (against gram-positive and gram-negative bacteria) with low cytotoxicity to skin/eukaryotic cells. [97]he delivered substances identified in the in vitro release studies were chlorogenic acid, caffeic acid, ferulic acid, p-coumaric acid, rosmarinic acid, and pinobanksin, Figure 5.The release profiles of all substances were similar (beginning of the release at 0.5 h and maximum release at 4h of testing, reaching a constant release/plateau from 4 h to 96 h). [98]he burst release is adequate for wound dressings since the delivery of the drug / active principle would prevent infection evolution.In pills, for example, it is expected that 75% of the active substance would be delivered within 45 minutes of administration. [99]Two release models were used to fit the results and the R2 values, Korsmeyer-Peppas (K-P) model (Equation 4) [100] and Peppas-Sahlin (P-S) model, [101] Figure 5. Peppas-Sahlin model is the best fit for the studied substances, but caffeic acid release was not properly adjusted by the studied models.All other substances were released according to the Peppas-Sahlin model, Equation 4. [101] This model consists of 2 terms on the right-hand side: the first term is the Fickian contribution, and the second term is the Case-II relaxational contribution. [102]A high k 1 value represents the drug diffusion release mechanism, while a high k 2 means polymer relaxation or heterogeneous erosion as the release mechanism. [103][104]Since negative k values should not be included in this analysis, [103] mainly non-linear Peppas-Sahlin propolis diffusion mechanism would be responsible for the release of the substances.

Conclusion
Flavonoids and phenolic acids were identified in the studied propolis extracts, where the main substance was kaempferide, an antimicrobial substance.There were also considerable amounts of rosmarinic acid, p-coumaric acid, ferulic acid, and pinobanksin.Propolis had antioxidant properties, identified through DPPH, FRAP, and ABTS and it was active against S. aureus, no matter the extraction method.Propolis was successfully incorporated into CMC gels.The swelling capacity of the gels might be dosedependent with the agent added, whereas CMC gels could be considered superabsorbent.The low swelling capacity and high gel fraction of CMC-P would be the consequence of propolis filling the pores of the membrane.Since the ethanolic extract of propolis is quite resinous, it may be responsible for the low absorption of moisture, leading to a low swelling degree and weight loss (biodegradability in water).Propolis could be anchoring the CMC chains, which was also observed by FTIR, where there was interaction and bonding between components.To corroborate the previous observation, propolis led to low CMC crystallite size formation (propolis could physically interfere with the CMC molecules, diminishing the possibility of contact between chains).The thermal degradation profile of CMC e CMC-P is similar, but the CMC-P sample degraded less than the CMC sample at temperatures above 400ºC.The CMC-P presented mainly a diffusion-controlled propolis release (Peppas-Sahlin model).The in vitro release studies showed a non-linear diffusion-based release kinetics for most phenolic substances of propolis extract (pinobankisin, rosmarinic acid, p-cumaric acid, ferulic acid, chlorogenic acid), characterizing a diffusion-controlled release system.The CMC-P samples present potential as a dressing material.

Figure 3 .
Figure 3. (a) FTIR spectra of CMC, propolis, and CMC-P samples, where new bands are indicated by arrows; XRD analysis of (b) CMC sample and (c) CMC-P sample, where fitted gauss curves ("peak") indicating the CMC diffraction peaks can be observed.

Figure 4 .
Figure 4. (a) Thermal degradation (TGA) curves of CMC and CMC-P samples and these curves' (1st) derivative curves; Samples (b) mechanical properties and (c) Young Modulus (E), elongation at break and failure strength.