alginate multilayer membranes as a system for controlled drug release in wound healing

Journal Pre-proof Silk fibroin/chitosan/alginate multilayer membranes as a system for controlled drug release in wound healing

Murilo Santos Pacheco, Gustavo Eiji Kano, Letícia de Almeida Paulo, Patrícia Santos Lopes, Mariana Agostini de Moraes PII:

S0141-8130(19)36979-X

DOI:

https://doi.org/10.1016/j.ijbiomac.2020.02.140

Reference:

BIOMAC 14757

To appear in:

International Journal of Biological Macromolecules

Received date:

29 August 2019

Revised date:

6 February 2020

Accepted date:

13 February 2020

Please cite this article as: M.S. Pacheco, G.E. Kano, L. de Almeida Paulo, et al., Silk fibroin/chitosan/alginate multilayer membranes as a system for controlled drug release in wound healing, International Journal of Biological Macromolecules(2018), https://doi.org/10.1016/j.ijbiomac.2020.02.140

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2018 Published by Elsevier.

Journal Pre-proof

Silk fibroin/chitosan/alginate multilayer membranes as a system for controlled drug release in wound healing Murilo Santos Pacheco1, Gustavo Eiji Kano1, Letícia de Almeida Paulo², Patrícia Santos Lopes2, Mariana Agostini de Moraes1* 1

Department of Chemical Engineering, Universidade Federal de São Paulo - UNIFESP,

Diadema, São Paulo, 09913-030, Brazil. 2

Department of Pharmaceutical Sciences, Universidade Federal de São Paulo - UNIFESP,

Diadema, São Paulo, 09913-030, Brazil. [email protected]

of

*

ro

Abstract

-p

In this study, we proposed the use of the biopolymers silk fibroin, chitosan and alginate, which are recognized for their biocompatibility and biodegradability, for the preparation

re

of multilayer membranes aiming at high performance wound dressings with controlled

lP

drug delivery. The rationale was to combine in one material the mechanical properties of fibroin, the antimicrobial action of chitosan and the ideal exudate absorption of

na

alginate, reaching a synergic effect of each biopolymer, without losing their individual intrinsic properties. The membranes were prepared by casting and diclofenac sodium was incorporated as model drug into the chitosan solution before the solvent

Jo ur

evaporation, being retained in the middle layer of the membrane. Morphological, thermal, mechanical, solubility and barrier properties of the membranes were evaluated, as well as cytotoxicity and microbiological permeation. Results show that the incorporation of the drug did not affect mechanical and barrier properties, as well as microbiological permeation. Drug release was evaluated in vitro using simulated solution of wound exudate at 37 °C and diclofenac sodium was released from the multilayer membrane in 7 hours, in which Fickian diffusion was the main mechanism associated. The results show the potential application of the biopolymer multilayer membranes as high-performance wound dressings. Keywords: biopolymers; casting; wound dressing; multilayer membrane; biomaterial.

1

Journal Pre-proof 1. Introduction Chronic wounds represent a great risk for human health, since they make easier the access of organisms that can cause infections and other diseases in the human body[1]. High-performance wound dressings can accelerate wound healing and allow the incorporation of drugs that will stay in direct contact with the injured region, controlling release and allowing a prolonged and effective action at the wound site[2]. Biopolymers have been used in the development of high-performance wound dressings, since they have properties that contribute to the tissue recovery process, such

of

as biocompatibility, biodegradability, microbial protection, exudate absorption, among others. In addition, they are environmentally more advantageous materials than

ro

synthetic polymers and enable controlled release of bioactive compounds[3]. The use of

-p

biopolymers can be studied in the form of membranes containing multiple layers, which offers versatility, ease of preparation, and drug incorporation capability.

re

The most commonly used technique for the construction of multilayer

lP

membranes is the so-called Layer-by-Layer (LbL), which is based on the electrostatic interaction between molecules of different charges for the formation of ultrathin layers[4]. Casting is another technique that has been studied to obtain multilayer

na

membranes and it is based on LbL since it uses the principle of the electrostatic interactions for the determination of the order of layers deposition. The use of

Jo ur

oppositely charged solutions for forming the layers leads to the interaction of the polyelectrolytes at the interface between the layers of the film[5]. This configuration also allows the slow release of drugs from the multilayer system, in which the dressing has a prolonged drug effect[6].

Some properties of multilayer membranes need to be considered for controlled release applications, such as matrix structure and properties, release kinetics and mass transfer mechanism involved in drug release. The study of the drug release profile allows a correlation with the real application conditions and allows to know if the matrix is suitable or not for the proposed functionality. In addition, the mechanism that controls release can be found from mathematical models applied to kinetic data. The two major release mechanisms are diffusion and swelling. The first is characterized by the movement of molecules caused by a concentration gradient and the absence of significant changes in the shape of the matrix during the process, while the second is 2

Journal Pre-proof characterized by the relaxation of the interactions between the biopolymer macromolecules and the increase of the system volume due to the contact of the external fluid. There are several possible mechanisms that could describe the ways that drug molecules are released from a matrix, such as diffusion, when there is the movement of drug molecules caused by a concentration gradient and the absence of significant changes in the shape of the matrix; swelling, when there is the water absorption by the matrix and the relaxation of the polymer chains; degradation, when there is the erosion, hydrolysis or dissolution of the matrix according to the medium conditions; and other reported like osmotic pumping and polymer-drug interactions [7–9]. For biopolymer

of

matrices, the release mechanism is even more complex, since biopolymers may be influence by the external conditions, especially for applications in the human body; e.g.,

ro

chitosan degradation by lysozyme [10,11] and alginate dissolution by Ca-Na ions

-p

exchanges [12], which could represent important drug release mechanisms in this context.

re

Silk fibroin (SF) is a protein extracted from silkworm cocoons that has excellent

lP

mechanical properties and it can be shaped in forms like films, scaffolds and hydrogels for drug release applications[6,13]. Chitosan (CS) is obtained from the deacetylation of chitin, a polysaccharide found in the exoskeleton of crustaceans and arthropods, and it

na

has characteristics such as excellent biocompatibility and antimicrobial activity[14]. Sodium alginate (SA) is a salt of alginic acid, a polysaccharide found in all species of

Jo ur

brown algae, and it can be widely used for wound dressings applications as it has swelling ability and assists in healing process by absorbing exudate and creating a moist environment[15].

In this context, we have developed multilayer membranes of silk fibroin, chitosan and alginate, containing a drug model, in order to develop a suitable material with potential application as a high-performance wound dressing with controlled drug release. For that, the physicochemical and biological characterization of the membranes was assessed, as well as the release evaluation of diclofenac sodium (which is a wellestablished model drug, with knowing quantification, solubility and diffusion parameters) and the elucidation of the mechanism of release through the application of mathematical models to kinetic data.

3

Journal Pre-proof 2. Materials and methods 2.1. Materials Chitosan (from shrimp shells, ≥75% deacetylated), alginic acid sodium salt (from brown algae Macrocystis pyrifera, medium viscosity) and diclofenac sodium were purchased from Sigma Aldrich (Brazil). Silkworm cocoons from Bombyx mori, for SF solution preparation, were supplied by silk company Bratac (Brazil). Glycerol was purchased from Synth (Brazil). 2.2. Methods

of

2.2.1. Preparation of solutions

ro

SA and CS aqueous solutions were prepared in concentrations of 2% (w/v). SA was dissolved in distilled water, while CS in 3% (v/v) acetic acid solution, both with the

re

using NaOH aqueous solution (1 mol/L).

-p

aid of a magnetic stirrer. CS solution was vacuum filtered, and pH was adjusted to 5.5

For sericin removal, silkworm cocoons were degummed with 0.1% (w/v)

lP

Na2CO3 solution three times for 30 minutes at 85 °C. The fibroin fibers were rinsed with abundant ultrapure water and dried at room temperature for 48 h and then

na

dissolved in CaCl2:CH3CH2OH:H2O (1:2:8 molar ratio) at 85°C for 90 minutes, to a final concentration of 5% (w/v). SF solution was dialyzed using a cellulose membrane

Jo ur

of 3.5 kDa molecular weight-cutoff (ThermoScientific) immersed in ultrapure water for 3 days at 8 °C to remove the salts of the solvent. The dialysis water was changed every 24 hours, keeping the ratio of solution to water 1:15. SF 2% (w/v) aqueous solution was obtained after the dialysis and it was used for membranes preparation. 2.2.2. Preparation of multilayer membranes The SF/CS/SA multilayer membranes were prepared by casting method in Petri dishes (Fig 1). Initially, glycerol (Gly), which acts as a plasticizer agent, was gently mixed with each individual biopolymer solution in the ratio of 1:1 in relation to the dry mass of biopolymer. SF+Gly solution was disposed in a Petri dish at the ratio of approximately 0.45 mL/cm² area and dried at 40 °C until about 90% of the total mass had been evaporated. After that, CS+Gly solution was added on the top of the partially dried layer of SF+Gly and dried at the same conditions until 80% of the total mass had been evaporated. Finally, SA+Gly solution was added over the pre-formed bilayer 4

Journal Pre-proof membrane and totally dried at the same conditions, forming a membrane composed of three layers: SF on the bottom, CS on the middle and SA on the top. To ensure a fair distribution of the layers regardless of the size of the Petri dish, it was disposed the same volume of each Biopolymer+Gly solution in a ratio of approximately 0.45 mL for 1 cm² of Petri dish area (i.e., for a 6 cm diameter Petri dish (28.3 cm²), we added 13 mL of SF 2% (w/v) solution containing 0.26 g of Gly). The same volume of solution and Gly weight were used for the other layers (CS+Gly and SA+Gly) and the 0.45 mL/cm² ratio was used when other sizes of membrane were needed. The membrane was placed in UV flow for irradiation for 30 min on each side as crosslinking method. For

of

biological tests, which requires the sterilization of the samples by UV irradiation, membranes were irradiated only once (30 min on each side for simultaneously

Jo ur

na

lP

re

-p

ro

crosslinking and sterilization).

Figure 1. Schematic illustration of multilayer membrane construction. SF = silk fibroin, CS = chitosan, SA = sodium alginate and Gly = glycerol.

2.2.3. Drug incorporation Drug incorporation was performed by adding the drug directly on CS solution, before solvent evaporation and layer formation. Thus, the diclofenac sodium (DS) was retained and dispersed in the biopolymer matrix of the intermediate layer (Fig. 2). A DS 5

Journal Pre-proof aqueous solution was prepared in a concentration of 0.5% (w/v) and added to the solution of CS+Gly in the proportion of 0.05 g DS to each 1.0 g dry mass of chitosan

-p

ro

of

using a magnetic stirrer.

re

Figure 2. Representative scheme of multilayer membrane with DS.

2.2.4.1.

lP

2.2.4. Characterization

Morphological analysis

na

The morphology of the membranes was analyzed by Scanning Electron Microscope (SEM) (model JSM-6610LV, JEOL) with acceleration voltage of 10 kV.

Jo ur

The samples were frozen with liquid nitrogen, fractured, placed on aluminum strips and then covered with a gold thin layer. The fracture of the membranes was analyzed in order to visualize the structure and delamination of the layers. The thickness of the layers was measured using ImageJ software. 2.2.4.2.

Thermal analysis

Thermogravimetric analysis (TGA) was obtained using the Shimadzu instrument with a DTG-60H model detector with a temperature range of 25 to 600 °C, at a heating rate of 10 °C/min and nitrogen flow of 50 mL/min. Multilayer membranes with and without incorporation of the drug were analyzed, as well as membranes of each one of the three biopolymers. The results obtained were normalized dividing the values of TGA and DTG by the weight of the sample analyzed. 2.2.4.3.

Solubility in water

6

Journal Pre-proof In order to evaluate the solubility aspects of the membranes, a water solubility test was performed. Samples of 3 cm x 1 cm were equilibrated at 50% relative humidity for 48 h and quantified. They were then immersed in Falcon tubes containing 5 mL of ultrapure water for 2 h, 6 h, 12 h, 24 h, 3 days and 7 days at 37 ° C. The samples were then dried for 24 h at 37 ° C, equilibrated at 50% relative humidity for 48 h and quantified. The weight loss in water was determined from Equation 1. The samples before and after the solubility test were also analyzed by SEM and TGA.

of

(1) Where wL is the weight loss in water (%), wi is the initial weight of the sample

Fluid Handling Capacity (FHC)

-p

2.2.4.4.

ro

(g) and wf is the weight of the sample at the end of the test (g).

The Fluid Handling Capacity (FHC) of the membranes with and without the

re

incorporated drug was determined following the British standard BS EM 13726-1 for

lP

hydrocolloids and dressings. The membranes were cut into circular dimensions of 4.5 cm in diameter, equilibrated at 50% relative humidity for 48 h and quantified. They

na

were then placed in the mouthpiece of a 3 cm diameter modified Paddington bottle containing 20 mL of simulated exudate fluid (SEF - 142 mmol/L of sodium ions and 2.5 mmol/L of calcium ions). The set (Paddington, SEF and membrane) was quantified and

Jo ur

inverted so that the SEF was in contact with the membrane. The inverted set was placed in desiccator containing silica gel (humidity below 20%) at 37 °C for 24 h. Finally, the whole set was removed from the desiccator, equilibrated at room temperature for 30 min and quantified again. FHC is the sum of fluid absorption (ABS) and moisture vapor transmission rate (MVTR) through the membrane, calculated by Equations 2 to 4. (2)

(3) (4)

7

Journal Pre-proof Where wmi is the weight of the membrane before the test, wmf is the weight of the membrane in the end of the test, wsi is the weight of the set before the test, wsf is the weight of the set in the end of the test, t is the time and A is the surface area. 2.2.4.5.

Mechanical properties

To evaluate the tensile strength of the membranes, the test was performed according to ASTM D882-10 (2010). Samples with 10 cm x 2.5 cm dimensions were equilibrated at 50% relative humidity for 48 h and placed in the texturometer CT3, by Ametek Brookfield, with the claws being 5 cm apart and the deformation rate adjusted

of

to 0.1 cm/s. From the data, it was possible to determine the tensile strength (TS) and the elongation at break (EB) of the membranes. Young's modulus was obtained using the

ro

tangent modulus values of the inclination of the tensile vs. deformation curves during

2.2.4.6.

Biological tests

-p

the elastic elongation phase.

re

The ability of membranes to prevent microbial penetration was tested using

lP

ionizing radiation-sterile PVC fittings embedded in glass vials containing 50 mL of TSB medium. The membrane samples were cut to diameters of 5 cm and sterilized for 30 min on each side by UV radiation. The vials with samples, which were inserted

na

between the PVC connection, were sealed with parafilm all around the connection. For the negative control, the connection as well as the sides were sealed and for the positive

conditions,

Jo ur

control no seals were made. The systems were exposed for 10 days under room being

macroscopically

evaluated

the

turbidity

and

growth

of

microorganisms on days 0, 5 and 10 of the experiment. For cytotoxicity tests, each film was cut (6 cm2/mL) and sterilized by exposure to UV light for 30 minutes on each side and then placed separately in sterile tubes. DMEM culture medium plus 10% fetal bovine serum (FBS) was added and the films remain for 24 h at 37 °C for preparation of the extracts, which was filtered with 0.45 μm pore membranes, and diluted with culture medium up to the concentration of 6.25% of the extract. In 96-well plates (each well with 0.28 cm2 area) were seeded Balb 3T3/c in the proportion of 2 x 104 cells per well with DMEM supplemented and then incubated for 24 h. After 24 h of sample addition, the culture medium was withdrawn and the wells were washed with the phosphate buffer solution (pH 7.4), then the Vital Neutral

8

Journal Pre-proof Red (NR) stain solution was added. After 3 h of incubation, the absorbance measurement at 540 nm for NR was carried out. 2.2.4.7.

Kinetics and mathematical modeling of drug release

Multilayer membranes with 6 cm in diameter were immersed in 50 mL of SEF at 37 °C under constant stirring. At pre-determined time intervals, the samples were transferred to another beaker containing SEF, at the same conditions, until reaching equilibrium. The concentration of the drug in each beaker used was determined by UVVis spectrophotometer (Thermo Scientific Genesys 10S) at the wavelength of 276 nm.

of

From the kinetic data obtained in the release tests, theoretical and empirical mathematical models were used to fit the experimental data in order to elucidate the

ro

release behavior and to predict the mechanisms involved in the release of the drug, such

-p

as diffusion, swelling or anomalous transport. The models used were Peppas (Equation 5), Peppas-Sahlin (Equation 6), Higuchi (Equation 7) and the simplified resolution of

re

the 2nd Fick’s Law, known as short time solution for plate geometry (Equation 8), and they can only be applied up to a fraction of 60% of the total mass of drug released[7,8].

lP

The OriginPro 8.0 software was used to fit the models.

Jo ur

na

(5)

(6)

√

√(

(7)

)

(8)

Where Mt/M∞ is the absolute cumulative amount of drug released, k is the kinetic constant, t is the time of release (s), n is the release exponent, k1, k2 and m are constants, DE is the diffusion coefficient for early time solution (cm²/s) and δ is the membrane thickness or diffusional distance (cm).

9

Journal Pre-proof 3. Results and discussion 3.1 Preparation of multilayer membranes The development of the multilayer membrane was tested with various configurations regarding the deposition order of the layers and the pH of the CS solution, evaluating the membrane stability in relation to macroscopic separation of the layers and the uniformity of the surface. When SF was tested as first layer followed by SA layer, layers delamination was observed, i.e., the layers were easily separated, probably due to the lack of electrostatic interactions among SA and SF. Also, if CS

of

solution was used at original pH (ca. 3.0), when in contact with the partially dried SF solution led to fibrillogenesis, in which the SF fibers precipitate, making the surface

ro

extremely rough, non-uniform, macroscopically phase separated and not ideal for the

-p

proposed application. The structure of better uniformity, without layers delamination or phase separation was that of SF as first layer, followed by CS with pH adjusted to 5.5

re

and finally SA as top layer.

lP

This structure allows the SF to be in contact with the external environment, providing mechanical support and barrier against microorganisms, the CS is in the middle layer, retaining the drug and offering antimicrobial activity, and the SA is in

na

contact with the wound, offering ideal moisture conditions for healing.

Jo ur

The multilayer membranes present a uniform structure and without macroscopic phase separation, with good cohesion between layers, caused by the electrostatic interactions between opposite charges of the biopolymers. SF is a protein with dominant negative charges at neutral aqueous solution (pI ~ 4.2)[16], while CS has positively charged amino groups (-NH3+, pKa ~ 6.3 to 7.0)[17] and SA has negatively charged carboxyl groups (-COO-, pKa ~ 3.38 to 3.65)[18]. Fig. 3 shows the SEM image of the membrane fracture with the thickness measured by ImageJ.

10

ro

of

Journal Pre-proof

re

-p

Figure 3. Micrograph of the multilayer membrane fracture surface. Thickness measurements of each layer were obtained by Image J software.

The multilayer membrane presented well delimited layers, but with varied

lP

thicknesses, which can be explained by the difference between the densities of the solutions once the same quantities of each solution were used[19]. CS solution

na

presented the major density, leading to a major layer between SF and SA layers.

Jo ur

3.2. Solubility in water

In order to analyze the stability of the multilayer membranes, the solubility of the membranes in water was analyzed at several time intervals up to 7 days. The results indicate a weight loss of 73.96% ± 1.89%. The high weight loss is possibly related to the entire dissolution of one of the layers, probably alginate, and to the dissolution of a great part of other layers (SF and CS), besides of solubilization of glycerin incorporated in all the three layers, since this plasticizer has high affinity and is completely miscible in water[20]. Thus, a crosslinking method based on exposure of the membranes to UV irradiation was proposed, since some studies show an improvement in the stability of biopolymer membranes after exposure to UV irradiation[21]. This is a physical crosslinking method, in which the UV irradiation generates free radicals in the polymer molecules, inducing the interactions between molecules chains and, thus, crosslinking the membranes[22–24]. After UV exposure, the average weight loss of the multilayer 11

Journal Pre-proof membranes was approximately 67.75% ± 2.72% for all time intervals analyzed in the solubility test, which represents a small gain with the crosslinking method. Fig. 4 shows SEM of multilayer membranes after 2 h of immersion in water, without and with exposure to UV irradiation. After UV irradiation, the third layer of the membrane is still present, which is not the case for the non-crosslinked membrane. In addition, all the membrane layers remained more intact, a fact confirmed by the

lP

re

-p

ro

of

thicknesses presented in each individual layer.

na

Figure 4. Micrographs of the multilayer membrane fracture surface after immersion in water for 2 hours. (a) membrane without any crosslinking and (b) membrane with UV crosslinking.

Jo ur

The thermal analysis confirms the maintenance of the multilayer configuration with UV crosslinking after immersion in water for 2 h. Fig. 5 shows that the peak formation of the second stage of the SA occurs at approximately 210 °C for the sample with UV crosslinking and the absence of that peak for the sample without crosslinking. This result confirms our hypothesis of SA layer dissolution when membranes were not crosslinked. The dissolution of this layer can be explained by the fact that SA is a watersoluble compound, owing to the presence of many -OH groups in its structure[25]. Since some controlled release systems use just the degradation/erosion of the matrix for the release of the active[26], solubility of the SA layer and possible partial solubility of the other layers can still represent a good characteristic of the membrane. Anyway, solubility test was performed under extreme conditions, in which the membrane was surrounded by water, leading to a maximization of dissolution effect of the layers compared to the real application condition (only one side of the membrane would be in contact with the aqueous medium). 12

na

lP

re

-p

ro

of

Journal Pre-proof

Jo ur

Figure 5. Thermograms of the multilayer membranes before immersion in water; and without and with UV crosslinking after 2 h of immersion in water. Inset: first derivative thermograms (DTG) of pure biopolymers SA, CS and SF.

3.3. Fluid Handling Capacity (FHC) The FHC of a dressing indicates the mass of fluid released by the wound capable of being drained by the dressing either by absorption or permeation in the form of moisture vapor. A skin wound, depending on its nature, can produce on average 5 g of exudate per 10 cm² of injured area per day[27,28]. Materials with high FHC values could lead to rapid dehydration of the wound by draining all the exudate produced, while low FHC values would cause excessive accumulation of exudate under the dressing, which could cause problems for healing, as a back pressure effect on the wound; maceration or wound infection; leakage from the edge of the dressing; and the movement of the dressing in accommodating the fluid; all of them causing discomfort, pain and/or breaking the microbiological barrier previously offered by the dressing [28,29]. 13

re

-p

ro

of

Journal Pre-proof

lP

Figure 6. FHC of membranes with and without DS incorporated. MVTR is the moisture vapor transmission rate and ABS is the absorption capacity.

na

The average values of FHC obtained for multilayer membranes without and with the incorporation of DS were, respectively, 2.39 ± 0.22 g/10 cm²/24 h and 2.55 ± 0.29

Jo ur

g/10 cm²/24 h (Fig 6). These results indicate an optimal fluid drainage capacity, since the ideal removal suggested by several authors represents 50% of the exudate produced, that is, FHC values on average of 2.5 g of exudate per 10 cm² per day[27,28,30]. This may be related to the high absorption capacity and the hydrophilic property of the SA[15], as well as the moisture retention capacity and water vapor permeation of SF and CS[31,32]. The values found are consistent when compared to the results of FHC reported in the literature for other types of membranes, such as the range of 3.79 to 4.20 g/10 cm²/24 h for amniotic membranes[27]. Estevam et al. (2012) found in their study a FHC result of 6.67 g/10 cm²/24 h for chitosan 2% (w/w) films[29], while Zhang et al. (2017) found a mean of 8.81 g/10 cm²/24 h for silk fibroin films in a thickness range of 40 to 300 µm[33], both of them significantly higher than the values we found for the multilayer membranes developed in the present work. which could be related to the membrane thickness and/or degree of crosslinking. In the first case [29], the authors did 14

Journal Pre-proof not report a crosslinking method for chitosan films; the absence or poor crosslinking could explain a great MVTR and absorption capacity of chitosan chains, leading to a high FHC value. In the other case [33], the authors found that the low crosslinking degree of SF films made them loose and random macromolecular structure, which helps the passage of water vapor, leading to high FHC values. Besides that, Zhang et al. (2017) also found that the thicker SF films, the lower the MVTR value[33], which could be related to a greater diffusion resistance by water vapor. Our FHC lower values compared with those studies could be explained by the high thickness of the membrane (> 300 µm) as well as by the fact that the multilayer membrane has three different

of

biopolymers with three different contributions to the water absorption and resistance to

ro

the water vapor diffusion.

In addition, the incorporation of the drug did not cause a significant change in

-p

the fluid drainage capacity, which can be related to a good dispersion and homogenous

re

incorporation of the DS in the middle layer of the multilayer membrane. 3.4. Mechanical properties

lP

Table 1 shows the results obtained by tensile test. The incorporation of the drug

the FHC results.

na

did not change the mechanical properties of multilayer membranes, corroborating with

Table 1. Mechanical properties of membranes with and without DS incorporated.

membrane Without DS With DS

Thickness

Tensile

Elongation at

Young’s

(mm)

Strength (MPa)

failure (%)

modulus (MPa)

0.51 ± 0.09*

3.05 ± 0.41*

45.14 ± 6.23*

20.73 ± 4.64*

0.51 ± 0.03*

2.63 ± 0.35*

50.28 ± 8.44*

17.84 ± 3.52*

Jo ur

Multilayer

* indicates that the averages presented in the same column are not statistically different. According to Xu et al. (2005), the results of TS and EB for a cross section of a chitosan film, made with 1% lactic acid solution and 4:1 chitosan:glycerol ratio, were about, respectively, 26 MPa and 33%[34]. Noishiki et al. (2002) found for TS and EB about 30 MPa and 1% for silk fibroin/cellulose (100/0) blend film composed by SF 2% (w/v)[35]. On the other hand, Thu et al. (2012) found in their study TS of 20.82 MPa and EB of 23.78% for alginate 6% (w/v) film containing ibuprofen, ethanol, glycerol 15

Journal Pre-proof and propylene glycol, with a thickness of 0.69 mm[36]. Comparing the values obtained in the mechanical test, the multilayer membrane showed TS below and EB above the values obtained from the literature for single layer films, which could be related to preparation procedure and plasticizer. The low values of tensile strength can be explained by the presence of plasticizers (Gly) in the layers of the biopolymers (1:1 for each biopolymer), since the addition of Gly causes an increase in competition for hydrogen bonds, reducing the interactions between the chains of biopolymers [37]. This allows the chains of the biopolymers to have a greater possibility of movement, reducing the stiffness of the membrane and increasing its flexibility (greater elongation

of

at break).

ro

The Young's modulus indicates the stiffness or strength of the material to elastic deformation. For SF/gelatin blend multilayer membranes, Mandal et al. (2009) found

-p

Young’s modulus in a range from approximately 800 to 4316.3 MPa, varying % of gelatin in the blend composition and the number of layers (one, three or five layers)[6].

re

SF/CS/SA multilayer membranes with and without DS showed low values of Young’s

lP

modulus (Table 1), indicating the ductile characteristic of the membranes, that is, a greater elastic deformation probably due the presence of Gly as plasticizer.

na

3.5. Biological tests

For microbial permeation test, the positive control showed turbidity of the

Jo ur

culture medium, indicating that the medium was capable of propitiating microbial growth. On the contrary, the negative control, which did not show turbidity of the culture medium, indicates that the medium was sterile, that is, any growth would occur due to the external permeation of microorganisms. For the multilayer membranes, the vials did not show turbidity during the period of exposure to the environment. Thus, the membranes acted effectively as a barrier, protecting the interior of the vials and preventing the permeation of microorganisms. This test indicates the good potential of membranes to be used as dressing because of their ability to prevent microbial permeation and consequently secondary wound infections[38]. Cytotoxicity tests were performed according to ISO 10993-5 for the evaluation of the cytotoxic potential of the multilayer membranes. According to the standard, the material developed must present a minimum of 70% of cellular viability for the proposed application[39]. Fig. 7 shows the results of cell viability of the membranes.

16

re

-p

ro

of

Journal Pre-proof

na

lP

Figure 7. Cell viability (%) as a function of the different extract concentrations from multilayer membranes composed only of SF, CS and SA, with addition of Gly in the three layers and with addition of Gly in the three layers and DS in the intermediate layer of CS.

The multilayer membrane does not exhibit cytotoxicity according to the ISO

Jo ur

standard. In contrast, the presence of DS in the extract without dilution (100% extract concentration) makes the membrane cytotoxic, which can be explained by the high concentration of DS in the extract and by the cytotoxic potential of the drug, which presents a low value of IC 50. Glycerol, that was used as a plasticizer agent, also leads to a cytotoxicity action, mainly due to the cells osmotic damage at the used concentration at 37 °C – assay temperature – as previous reported by Davidson et al. (2015) [40]. The authors also reported the glycerol cytotoxicity to adherent cells as dose-dependent, corroborating the results presented at Figure 7. On the other hand, Wiebe & Dinsdale (1991) reported that glycerol effect could be reversible after it is removed, suggesting that the caused effect is more related to proliferation inhibition than a cytotoxicity effect per se [41]. Studies using SF films and CS/SF blends have reported results of noncytotoxicity when applied to fibroblasts, which is in accordance to the results obtained from the tests performed with the Balb 3T3 cells[42,43]. Thus, the observed 17

Journal Pre-proof cytotoxicity does not inhibit the application of the developed matrix, besides proving that the incorporated drug is released and that its activity was not compromised by the casting method used in the production of the membranes. 3.6. Kinetics and mathematical modeling of drug release Fig. 8 shows the DS release profile of the multilayer membranes for 420 min of experiment. There was an incomplete release, since about 70% of the incorporated drug was released to the SEF medium, which can be related to the lower solubility of DS in SEF medium when compared to water solubility. In addition, other studies report DS

of

release in CS/SF blends in the range of 65-75%, due to possible small amounts of drug

Jo ur

na

lP

re

-p

ro

particles not dissolved in solution and, therefore, not released[44].

Figure 8. Kinetic profile of DS released from multilayer membranes.

The release profile indicates that half of drug incorporated into the membrane is released in approximately 100 min, and that the equilibrium begins to be reached in approximately 400 min. Lopes et al. (2018) found in their study of chitosan-xanthan polyelectrolyte membranes loaded with indomethacin an equilibrium time of drug release around 300 min in PBS medium[45]. Rujiravanit et al. (2003) studied CS/SF blend films in which DS released equilibrium in several media of different pH were around 60 min[44]. Kulkarni et al. (1999) found equilibrium time of DS release from

18

Journal Pre-proof alginate beads crosslinked with glutaraldehyde in a range of 60-300 min, depending on time and temperature of crosslinking [46]. This indicates that the use of multilayer membrane offered greater resistance to diffusion, making the release of DS slower and more controlled than literature release data found for the same biopolymers. Fig. 9 shows the fitting of the models (Equations 3 to 6) to the release profile following the condition of complete renewal of the release medium and restriction of fitting data of up to 60% of the total mass released[8], while Table 2 shows the

na

lP

re

-p

ro

of

parameters obtained by fitting the data.

Figure 9. Fitting of mathematical models to the kinetic data of DS release from multilayer membranes.

Jo ur

Table 2. Parameters of the mathematical models fitted to the kinetic release data.

Model

Parameters

R²

k

0.093 ± 0.003

n

0.479 ± 0.008

k1

0.093 ± 0.003

k2

1.77E-19 ± 0.00

m

0.479 ± 0.008

Higuchi

k

0.087 ± 0.001

0.995

Early time solution

D (cm²/s)

3.63E-8 ± 4.58E-10

0.995

Peppas

Peppas-Sahlin

0.996

0.996

19

Journal Pre-proof All applied models provided good fitting to the release kinetic data, with correlation coefficients of 0.99. The Peppas and Peppas-Sahlin models presented the highest R², that is, the models that best describe the release of DS from the multilayer membrane. Thus, from the coefficient ‘n’ of the Peppas model it is possible to determine the rate-controlling mechanism involved in the release of the drug. Considering a film of small thickness, when n = 0.5, the release mechanism is controlled by diffusion and transport is known as Fickian. For n = 1.0, the mechanism is controlled by the swelling of the polymer, being case-II transport. For the case where 0.5 < n < 1.0, transport is considered anomalous, where diffusion and swelling

of

mechanisms overlap[7]. Our coefficient ‘n’ is very close to 0.5, and it can be determined that Fickian diffusion is the mass transfer mechanism that dominates the

ro

release rate in the biopolymer multilayer system. This is also evidenced by the fitting of

-p

the Peppas-Sahlin model, where k1 resulted in a much larger value than k2, that is, the term referring to the influence of matrix swellability is negligible with respect to the

re

contribution of the term referring to diffusion.

lP

Similar Fickian diffusion behaviors have been reported in studies involving release of Trypan Blue dye in SF membranes[6]. In contrast, studies have been carried out using SA particles for the release of DS and ibuprofen, and report anomalous

na

transport behavior, that is, overlapping diffusion and swelling mechanisms[49].

Jo ur

The mathematical models usually used for drug release are relatively simple and only consider Fickian, case-II transport and anomalous mechanisms, but there are other factors that can act in drug release, such as drug solubility and partition [50], pH of the medium and biopolymers dissolution (matrix erosion). As DS is a water-soluble drug, derived from a weak acid (pKa = 4.0) [51], its release is favored in the medium used (SEF) because of the neutral pH. In addition, in our release test, the SEF medium was renewed with each concentration measure, forcing a drug concentration gradient between the matrix and the SEF, so drug partition may not have a great influence in this case. As erosive mechanisms can coexist with diffusive ones for swellable matrices, it is important to consider biopolymers dissolution in this case. Although the multilayer membrane structure proposes the SA as the direct contact layer with the wound and the SF layer as external to the environment, the experimental method performed for drug release considers the release on both sides of the matrix, once the membrane is completely immersed in the SEF medium. Thus, the dissolution of the SA layer may be 20

Journal Pre-proof occurring very rapidly in relation to the total release time, even after treatment of the matrix with UV irradiation. Thus, with the rapid dissolution of the SA layer, that could be responsible for an anomalous behavior, we only have the maintenance of the CS and SF layers in the membrane, being responsible for the major behavior of Fickian diffusion of the drug during the release. In fact, when comparing the matrix before and after the test, erosion of one of the layers during the experiment occurs, but the integrity of the matrix is observed. This is probably due to the greater water stability of SF layer after UV crosslinking, besides the greater stability of the CS layer conferred by the addition of NaOH in the pH adjustment.

of

The fitting for the short-time model, or the semi-infinite solid model, also

ro

indicates that mass transfer is controlled by diffusion, since this model is a simplification of the resolution of 2nd Fick's Law of diffusion[8]. From this model, it

-p

was used δ = 0.384 ± 0.065 mm as average membrane thickness to calculate the diffusion coefficient of the drug in the membrane. The value obtained D = 3.63 x 10-8

re

cm²/s, indicates a lower diffusion than that found by Mandal et al. (2009) for different

lP

compounds released in silk fibroin/gelatin blend multilayered films[6] in a range from 2.43 x 10-7 to 2.81 x 10-7 cm²/s. The ‘D’ value obtained by fitting the model refers to an apparent diffusion coefficient of the full multilayer membrane (bulk), since the kinetic

na

data refers to concentrations of DS in solution, i.e., the drug that has already spread throughout all of the three biopolymers in the matrix until it reached the matrix-solution

Jo ur

interface and migrated into the SEF. 4. Conclusion

The production of SF/CS/SA multilayer membranes by casting method was successfully achieved. The multilayer membrane configuration brings alginate as the layer in contact with the wound, being absorbent and promoting healing, chitosan as intermediate layer, which has antimicrobial activity and host the drug and, in the outer layer, fibroin represents the wound protection to the external environment, bringing resistance and mechanical support to the dressing. The water solubility analysis showed the dissolution of the SA layer, which raised the need to expose the membranes to the UV irradiation process as a crosslinking method, resulting in an improvement in membrane stability after 2 hours of immersion in water. The DS incorporation did not change thermal, mechanical and barrier properties of the membranes. Drug release 21

Journal Pre-proof showed a slower and controlled release for multilayer membranes and that the Fickian diffusion behavior is the dominant release mechanism. The multilayer membranes produced showed great potential for application as high-performance wound dressings with important social and technological bias, according to the results obtained from the physicochemical, biological and barrier properties, besides the kinetics of drug release and mechanism of diffusion. Acknowledgments The authors would like to thank the São Paulo Research Foundation (FAPESP

of

#2017/20274-9 and #2016/17555-3).

A.J. Whittam, Z.N. Maan, D. Duscher, V.W. Wong, J.A. Barrera, M. Januszyk,

-p

[1]

ro

References

G.C. Gurtner, Challenges and Opportunities in Drug Delivery for Wound

J. Choi, T. Konno, M. Takai, K. Ishihara, Controlled drug release from

lP

[2]

re

Healing, Adv. Wound Care. 5 (2016) 79–88.

multilayered phospholipid polymer hydrogel on titanium alloy surface,

[3]

na

Biomaterials. 30 (2009) 5201–5208.

C.H. Goh, P. Wan, S. Heng, L.W. Chan, Alginates as a useful natural polymer

Jo ur

for microencapsulation and therapeutic applications, Carbohydr. Polym. 88 (2012) 1–12. [4]

G. Decher, J.D. Hong, J. Schmitt, Buildup of ultrathin multilayer films by a selfassembly process: III. Consecutively alternating adsorption of anionic and cationic polyelectrolytes on charged surfaces, Thin Solid Films. 210–211 (1992) 831–835.

[5]

L. Ma, W. Yu, X. Ma, Preparation and Characterization of Novel Sodium Alginate/Chitosan Two Ply Composite Membranes, J. Appl. Polym. Sci. 106 (2007) 394–399.

[6]

B.B. Mandal, J.K. Mann, S.C. Kundu, Silk fibroin/gelatin multilayered films as a model system for controlled drug release, Eur. J. Pharm. Sci. 37 (2009) 160–171.

[7]

J. Siepmann, N.A. Peppas, Modeling of drug release from delivery systems based on hydroxypropyl methylcellulose (HPMC), Adv. Drug Deliv. Rev. 48 (2001) 22

Journal Pre-proof 139–157. [8]

L. Serra, J. Doménech, N.A. Peppas, Drug transport mechanisms and release kinetics from molecularly designed poly(acrylic acid-g-ethylene glycol) hydrogels, Biomaterials. 27 (2006) 5440–5451.

[9]

S. Fredenberg, M. Wahlgren, M. Reslow, A. Axelsson, The mechanisms of drug release in poly(lactic-co-glycolic acid)-based drug delivery systems - A review, Int. J. Pharm. 415 (2011) 34–52.

[10] Y.M. Yang, W. Hu, X.D. Wang, X.S. Gu, The controlling biodegradation of

of

chitosan fibers by N-acetylation in vitro and in vivo, J. Mater. Sci. Mater. Med.

ro

18 (2007) 2117–2121.

[11] S. Aiba, Studies on chitosan: 4. Lysozymic hydrolysis of partially N-acetylated

-p

chitosans, Int. J. Biol. Macromol. 14 (1992) 225–228.

re

[12] A. Kikuchi, M. Kawabuchi, A. Watanabe, M. Sugihara, Y. Sakurai, T. Okano, Effect of Ca2+-alginate gel dissolution on release of dextran with different

lP

molecular weights, J. Control. Release. 58 (1999) 21–28. [13] P.-P.Q. Qiao-Na Wei, Ai-Min Huang, Lin Ma, Zilun Huang, Xi Huang, L.Z.

na

Zhu-Ping Gong, Structure Regulation of Silk Fibroin Films for Controlled Drug Release, J. Appl. Polym. Sci. 125 (2012) 477–484.

Jo ur

[14] Q. Wang, Y.M. Du, L.H. Fan, Properties of chitosan/poly(vinyl alcohol) films for drug-controlled release, J. Appl. Polym. Sci. 96 (2005) 808–813. [15] K.Y. Lee, D.J. Mooney, Alginate: Properties and biomedical applications, Prog. Polym. Sci. 37 (2012) 106–126. [16] Q. Lu, H. Zhu, C. Zhang, F. Zhang, B. Zhang, D.L. Kaplan, Silk self-assembly mechanisms and control from thermodynamics to kinetics, Biomacromolecules. 13 (2012) 826–832. [17] A.G. Karakeçili, C. Satriano, M. Gümüsderelioglu, G. Marletta, Surface Characteristics of Ionically Crosslinked Chitosan Membranes, J. Appl. Polym. Sci. 106 (2007) 3884–3888. [18] F.A. Simsek-Ege, G.M. Bond, J. Stringer, Polyelectrolye complex formation between alginate and Chitosan as a function of pH, J. Appl. Polym. Sci. 88

23

Journal Pre-proof (2003) 346–351. [19] R.G. de Paiva, M.A. de Moraes, F.C. de Godoi, M.M. Beppu, Multilayer Biopolymer Membranes Containing Copper for Antibacterial Applications, J. Appl. Polym. Sci. 126 (2012) 18–24. [20] Y. Ma, L. Xin, H. Tan, M. Fan, J. Li, Y. Jia, Z. Ling, Y. Chen, X. Hu, Chitosan membrane dressings toughened by glycerol to load antibacterial drugs for wound healing, Mater. Sci. Eng. C. 81 (2017) 522–531. [21] M.C.F.C. Felinto, D.F. Parra, C.C. da Silva, J. Angerami, M.J.A. Oliveira, A.B.

of

Lugão, The swelling behavior of chitosan hydrogels membranes obtained by UV-

ro

and γ-radiation, Nucl. Instruments Methods Phys. Res. B. 265 (2007) 418–424. [22] M.S. McCaig, D.R. Paul, Effect of UV crosslinking and physical aging on the gas

-p

permeability of thin glassy polyarylate films, Polymer (Guildf). 40 (1999) 7209–

re

7225.

[23] N. Reddy, R. Reddy, Q. Jiang, Crosslinking biopolymers for biomedical

lP

applications, Trends Biotechnol. 33 (2015) 362–369. [24] M.P. Ohan, K.S. Weadock, M.G. Dunn, Synergistic effects of glucose and

na

ultraviolet irradiation on the physical properties of collagen, J. Biomed. Mater. Res. 60 (2002) 384–391.

Jo ur

[25] A.D. Augst, H.J. Kong, D.J. Mooney, Alginate Hydrogels as Biomaterials, Macromol. Biosci. 6 (2006) 623–633. [26] B. Singh, J.S. Kanwar, P. Kumari, Modification of Dietary Fiber Psyllium with Poly ( vinyl pyrrolidone ) through Network Formation for Use in Slow Drug Delivery Application 1, Polym. Sci. 60 (2018) 331–348. [27] R. Singh, M.P. Chacharkar, Dried gamma-irradiated amniotic membrane as dressing in burn wound care, J. Tissue Viability. 20 (2011) 49–54. [28] D. Queen, J.D.. Gaylor, J.. Evans, J.. Courtney, W.. Reid, The preclinical evaluation of the water vapour transmission rate through burn wound dressings, Biomaterials. 8 (1987) 367–371. [29] L.S. Estevam, H.S. Debone, C.M.P. Yoshida, C.F. Da Silva, Adsorption of bovine serum and bovine haemoglobin onto chitosan film, Adsorpt. Sci. Technol.

24

Journal Pre-proof 30 (2012) 785–792. [30] L. Atkin, J. Stephenson, S.D. Bateman, Foam dressings: A review of the literature and evaluation of fluid-handling capacity of four leading foam dressings, Wounds UK. 11 (2015) 75–81. [31] S. Despond, E. Espuche, A. Domard, Water Sorption and Permeation in Chitosan Films: Relation between Gas Permeability and Relative Humidity, J. Polym. Sci. Part B Polym. Phys. 39 (2001) 3114–3127. [32] H. Kweon, H.C. Ha, I.C. Um, Y.H. Park, Physical properties of silk

of

fibroin/chitosan blend films, J. Appl. Polym. Sci. 80 (2001) 928–934.

ro

[33] W. Zhang, L. Chen, J. Chen, L. Wang, X. Gui, J. Ran, G. Xu, H. Zhao, M. Zeng, J. Ji, L. Qian, J. Zhou, H. Ouyang, X. Zou, Silk Fibroin Biomaterial Shows Safe

-p

and Effective Wound Healing in Animal Models and a Randomized Controlled Clinical Trial, Adv. Healthc. Mater. 6 (2017) 1–16.

re

[34] Y.X. Xu, K.M. Kim, M.A. Hanna, D. Nag, Chitosan – starch composite film :

lP

preparation and characterization, Ind. Crops Prod. 21 (2005) 185–192. [35] Y. Noishiki, Y. Nishiyama, M. Wada, S. Kuga, J. Magoshi, Mechanical

na

properties of silk fibroin-microcrystalline cellulose composite films, J. Appl. Polym. Sci. 86 (2002) 3425–3429.

Jo ur

[36] H.E. Thu, M.H. Zulfakar, S.F. Ng, Alginate based bilayer hydrocolloid films as potential slow-release modern wound dressing, Int. J. Pharm. 434 (2012) 375– 383.

[37] M.A. Bertuzzi, Mechanical properties of a high amylose content corn starch based film , gelatinized at low temperature, Brazilian J. Food Technol. 15 (2012) 219–227. [38] S. Wittaya-areekul, C. Prahsarn, Development and in vitro evaluation of chitosan – polysaccharides composite wound dressings, Int. J. Pharm. 313 (2006) 123– 128. [39] I. Standard, ISO 10993-5:2009 Biological evaluation of medical devices -- Part 5: Tests for in vitro cytotoxicity, Int. Organ. Stand. (2009). [40] A.F. Davidson, C. Glasscock, D.R. McClanahan, J.D. Benson, A.Z. Higgins,

25

Journal Pre-proof Toxicity Minimized Cryoprotectant Addition and Removal Procedures for Adherent Endothelial Cells, PLoS One. 10 (2015) 1–22. [41] J.P. Wiebe, C.J. Dinsdale, Inhibition of cell proliferation by glycerol, Life Sci. 48 (1991) 1511–1517. [42] T. Liu, J. Miao, W. Sheng, Y. Xie, Cytocompatibility of regenerated silk fibroin film : a medical biomaterial applicable to wound healing, J. Zhejiang Univ. 11 (2010) 10–16. [43] W. Luangbudnark, J. Viyoch, W. Laupattarakasem, P. Surakunprapha, P.

of

Laupattarakasem, Properties and biocompatibility of chitosan and silk fibroin blend films for application in skin tissue engineering, Sci. World J. 2012 (2012)

ro

1–10.

-p

[44] R. Rujiravanit, S. Kruaykitanon, A.M. Jamieson, S. Tokura, Preparation of Crosslinked Chitosan / Silk Fibroin Blend Films for Drug Delivery System,

re

Macromol. Biosci. 3 (2003) 604–611.

lP

[45] S.A. Lopes, I.G. Veiga, A.C. Krause, A. Luiza, R. Pires, Â.M. Moraes, M. Moraes, Physicochemical properties and release behavior of indomethacin-loaded

na

polysaccharide membranes, Int. J. Polym. Mater. Polym. Biomater. 0 (2018) 1–9. [46] A.R. Kulkarni, K.S. Soppimath, T.M. Aminabhavi, Controlled release of

Jo ur

diclofenac sodium from sodium alginate beads crosslinked with glutaraldehyde, Pharm. Acta Helv. 74 (1999) 29–36. [47] S.A. Lopes, I.G. Veiga, A.C. Krause, A. Luiza, R. Pires, Â.M. Moraes, M. Moraes, Physicochemical properties and release behavior of indomethacin-loaded polysaccharide membranes, Int. J. Polym. Mater. Polym. Biomater. 0 (2018) 1–9. [48] S.M. Samani, H. Montaseri, A. Kazemi, The effect of polymer blends on release profiles of diclofenac sodium from matrices, Eur. J. Pharm. Biopharm. 55 (2003) 351–355. [49] K.S. V Krishna Rao, M.C.S. Subha, B. Vijaya Kumar Naidu, M. Sairam, N.N. Mallikarjuna, T.M. Aminabhavi, Controlled release of diclofenac sodium and ibuprofen through beads of sodium alginate and hydroxy ethyl cellulose blends, J. Appl. Polym. Sci. 102 (2006) 5708–5718. [50] R. Bettini, P.L. Catellani, P. Santi, G. Massimo, N.A. Peppas, P. Colombo, 26

Journal Pre-proof Translocation of drug particles in HPMC matrix gel layer: Effect of drug solubility and influence on release rate, J. Control. Release. 70 (2001) 383–391. [51] L. Yang, R. Fassihi, Modulation of diclofenac release from a totally soluble controlled release drug delivery system, J. Control. Release. 44 (1997) 135–140. [52] G.M. Nogueira, M.A. de Moraes, A.C.D. Rodas, O.Z. Higa, M.M. Beppu, Hydrogels from silk fibroin metastable solution: Formation and characterization

Jo ur

na

lP

re

-p

ro

of

from a biomaterial perspective, Mater. Sci. Eng. C. 31 (2011) 997–1001.

27

Journal Pre-proof Silk fibroin/chitosan/alginate multilayer membranes as a system for controlled drug release in wound healing Murilo Santos Pacheco1, Gustavo Eiji Kano1, Letícia de Almeida Paulo², Patrícia Santos Lopes2, Mariana Agostini de Moraes1* 1

Department of Chemical Engineering, Universidade Federal de São Paulo - UNIFESP,

Diadema, São Paulo, 09913-030, Brazil. 2

Department of Pharmaceutical Sciences, Universidade Federal de São Paulo -

UNIFESP, Diadema, São Paulo, 09913-030, Brazil.

of

[email protected]

ro

*

-p

Author statement

Murilo Santos Pacheco: Investigation, data curation, formal analysis, methodology,

re

writing – original draft

lP

Gustavo Eiji Kano: Investigation, data curation, formal analysis Letícia de Almeida Paulo: Investigation

na

Patricia Santos Lopes: formal analysis, writing – review and editing Mariana Agostini de Moraes: Supervision, Project administration, funding acquisition,

Jo ur

writing – review and editing

28