Controlled Release of Metformin Hydrochloride from Core-Shell Nanofibers with Fish Sarcoplasmic Protein

Background and Objectives: A coaxial electrospinning technique was used to produce core/shell nanofibers of a polylactic acid (PLA) as a shell and a polyvinyl alcohol (PVA) containing metformin hydrochloride (MH) as a core. Materials and Methods: Fish sarcoplasmic protein (FSP) was extracted from fresh bonito and incorporated into nanofiber at various concentrations to investigate the influence on properties of the coaxial nanofibers. The morphology, chemical structure and thermal properties of the nanofibers were studied. Results: The results show that uniform and bead-free structured nanofibers with diameters ranging from 621 nm to 681 nm were obtained. A differential scanning calorimetry (DSC) analysis shows that FSP had a reducing effect on the crystallinity of the nanofibers. Furthermore, the drug release profile of electrospun fibers was analyzed using the spectrophotometric method. Conclusions: The nanofibers showed prolonged and sustained release and the first order kinetic seems to be more suitable to describe the release. MTT assay suggests that the produced drug and protein loaded coaxial nanofibers are non-toxic and enhance cell attachment. Thus, these results demonstrate that the produced nanofibers had the potential to be used for diabetic wound healing applications.

examine drug release properties of the coaxial nanofibers. The produced nanofibers were characterized by scanning electron microscopy (SEM), fourier-transform infrared spectroscopy (FTIR) and differential scanning calorimetry (DSC). The in vitro drug release profile of nanofibers was evaluated by UV-VIS spectroscopy and cell viability examined via MTT assay.

Sarcoplasmic Protein (SP) Preparation
Fish sarcoplasmic protein (FSP) was extracted from fresh bonito by using the method presented in our previous study [24]. The bonitos stored at −30 • C were first thawed and chopped. They were homogenized with deionized water and then centrifuged for 15 min, 18,000 × g, 5 • C. After the centrifugation, the supernatant was freeze dried (Alpha 1-2 LD Plus; Christ, Osterode am Harz, Germany) for obtain the FSP powder. This study was carried out under protocol number FEN-B-121218-0614 obtained on 15 June 2018.

Preparation of Spinning Solutions
PLA solution was prepared at a concentration of 8% (w/v) in DCM:DMF (4:1 v/v) and stirred for about 1 h to obtain viscous solution. PVA solutions at 5% (w/v) of polymer content were prepared in distilled water and stirred at 80 • C for 4 h to ensure complete dissolution. Then, the solution was cooled down to room temperature, 1 wt%, 3 wt% and 5 wt% FSP contents and 2% (w/v) metformin hyrochloride (MH) (20 mg) were individually added into the PVA solution under constant stirring. Contents of the core and the shell spinning solutions of the samples is given in Table 1.

Coaxial Electrospinning
The nanofibers were produced by using a laboratory scale electrospinning unit (NS24, Inovenso Co., Istanbul, Turkey). Two individual syringes containing the core (PVA solution) and the shell (PCL solution) were connected to a coaxial spinneret and driven by two separate pumps. A voltage of 13 kV was supplied by a high-voltage power supplier during coaxial electrospinning. The distance between the coaxial nozzle and the grounded collector was set as 12 cm, the flow rate of PLA solution was 1.4 mL/h, and the PVA solution was 0.5 mL/h. All of the electrospinning processes were carried out under ambient conditions (25 • C with a relative humidity of 55%).

Protein Composition of FSP
The protein composition of the freeze-dried fish sarcoplasmic protein (FD FSP) was analyzed by a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) method using 5% (w/v) stacking gel and 12% (w/v) resolving gel according to the method of Laemmli (1970). An amount of 27.5 mg FD FSP sample was dissolved in 0.5 mL of demineralised water. It was then centrifuged and after centrifugation, the supernatant was used to determine the protein composition. The electrophoresis procedure was performed at 100 V for 15 min and at 120 V for 75 min by adjusting to pH 8.3 in tris glycine buffer. Samples were boiled for the denaturation procedure before the loading onto the gel. After the electrophoresis, proteins in the separating gel were determined by coomassie staining. Fermentas Protein Molecular Weight Marker (Pierce™ Unstained Protein MW Marker, Thermo Fisher Scientific, Waltham, MA, USA) containing seven proteins within 14.4-116 kDa range was used in order to determine the molecular weight range of the proteins by the Image Analyzer System (Chemi Doc MP Imaging System-BioRad, Hercules, CA, USA).

Scanning Electron Microscopy (SEM)
The morphologies of the electrospun fibers were examined using a scanning electron microscope(SEM) (EVA MA 10, Zeiss, San Diego, CA, USA) at an accelerating voltage of 10 kV. Samples were coated with gold/palladium (18 nm) using a sputter coater (SC7620, Quorum, Lewes, UK). The average fiber diameters and their distribution were determined by the image analysis software (SmartSEM, Zeiss, San Diego, CA, USA) using the measurement of 100 fibers selected randomly from each sample.

Fourier-Transform Infrared Spectroscopy
To determine the composition and possible interaction between the fibers components, the infrared spectrums of the fibers were registered using a fourier-transform infrared spectroscopy (FT-IR) (FT/IR 4700, Jasco, Tokyo, Japan) equipped with a Gladi attenuated total reflection (ATR) viewing plate (Diamond ATR crystal) and a liquid-nitrogen cooled mercury cadmium telluride (MCT) detector at a wavelength of 500 cm −1 to 4000 cm −1 . The nanofibers used in the drug release study for 21 days were dried at vacuum oven at 40 • C for one day before the FTIR test.

MTT Assay
The hacat cell lines, which is human keratinocyte cell line purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) were plated separately in tissue culture flasks and cultured in Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS). The cells were maintained in a humidified incubator at 37 • C with 5% CO 2 and they were trypsinized and counted with a hemocytometer. All the nanofibers were cut to a size of 1 cm 2 and sterilized in UV. DMEM cell culture medium containing 10% FBS, 1% Penicillin/Streptomycin as well as 100,000 cells were then seeded onto nanofibrous membranes in 24-well culture plate.
MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) was dissolved in PBS at 5 mg/ml. The stock MTT solution (66 µl per well) was added on the cells and materials in DMEM supplemented with 10% FBS, 1% Penicillin/Streptomycin in the culture plate incubated for 2 h at 37 • C and 5% CO 2 . After incubation, purple colored formazan crystal was formed at the bottom of the well and the above DMEM was withdrawn. DMSO solution was added to wells and measured for light absorbance in 96 wells/plate at 570 nm.

In Vitro Release Studies
In vitro drug release was performed by the metformin hydrochloride which was loaded into the PVA/PLA electrospun nanofibers. The drug loaded nanofibers were cut into 2 × 2 cm pieces, weighted and immersed in glass bottles containing 7 mL of phosphate buffer solution (PBS) and the pH of the PBS solution was adjusted to 7.4 then shaked at 37 • C for 21 days. At determined time intervals, the amount of drug relase was measured at 233 nm by UV-VIS spectrophotometer. Each experiment was repeated three times.

In Vitro Release Kinetics
To disclose the drug release mechanism, the MH release profiles were fitted to four popular mathematical equations such as zero order, first order, Higuchi and Peppas-Korsemeyer equations represented as Equations (1)-(4).
Zero order kinetic defines a linear relationship between the fractions of drug release versus time: where Q is the fractional amount of drug release at time t and "K 0 " is the zero-order release constant. The equation that describes first order kinetic is where Q is the fractional amount of drug release at time t and K 1 is the first-order release constant. Higuchi kinetic defines a linear dependence of the active fraction released per unit of square root of time: where Q is the fractional amount of drug release at time t and K h is the Higuchi kinetic constant. The equation of Korsmeyer-Peppas is where Q is the fractional amount of drug release at time t, K is the kinetic constant and n is the diffusion exponent which is indicative of the drug release mechanism.

Thermal Analysis
The thermal properties of the nanofibers were analyzed using a Perkin Elmer Jade differential scanning calorimeter (DSC). Samples were weighed and placed in alumina pans. The heating was performed from room temperature to 300 • C with a heating rate of 10 • C/min under dynamic argon atmosphere (20 mL/min).

SDS-PAGE Assay
SDS-PAGE (sodium dodecyl sulphate-polyacrylamide gel electrophoresis technique) is commonly used technique for characterization of proteins. Molecular weight distrubition of FSP were determined. As illustrated in Figure 1, with molecular weights (MW) ranging from 23 to 94 kDa were observed and it was identified that the 3840 and 48 kDa patterns were the most intense ones. These observations are mainly consistent with other studies [25,26]

Morphological Characterization of Nanofibers
SEM images and associated diameter histograms of the prepared nanofibers having different concentrations of FSP are shown in Figure 2. The core/shell nanofibers possess a smooth surface, which indicates a successful encapsulation of FSP and MH without deteriorating the morphology of the fibers. It was observed from the SEM images that all samples are defect-free, beadless, and geometrically uniform nanofibers. However, fiber diameter and fiber distribution were changed by the addition of protein and increasing its concentration. As shown in Figure 2, average diameters of MH, MHFSP1, MHFSP3, MHFSP5 were recorded as 647 nm, 621 nm, 623 nm, and 681 nm, respectively. The average diameter were affected by voltage value. Therefore, the constant voltage applied during nanofiber formation to examine the only FSP content affected the diameters [27,28]. Clearly, the addition of FSP decreased the average fiber diameters and the standard deviation. This can be attributed to the higher electrical conductivity of FSP, which caused much more charge over the ejected jet and thus, higher elongation. Furthermore, decreasing the standart deviation by the amount of FSP was due to highest chain entanglements in the core. The observed increase in the diameters of the nanofibers with an increase in the FSP content from %1 to %5 can be associated with the knowledge that an increase in solution concentration usually results in a larger fiber diameter. These observations are consistent with those of previous studies performed with different globular proteins [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]

Fourier-Transform Infrared Spectroscopy (FTIR)
FTIR spectroscopic imaging in ATR (Attenuated Total Reflection) mode permits investigation of the molecular and structural properties of polymers [31]. FTIR spectra of MH, MHFSP1, MHFSP3, MHFSP5 core-shell fibers are depicted in Figure 3. The spectra of nanofibers show the general PLA bands, including the CH 3 asymmetric stretching at 2948.6 cm −1 , C=O stretching at 1748.2 cm −1 and C-0=0 stretching at 1181.2 cm −1 and 1038.8 cm −1 , which is a characteristic ester bond. Bending vibrations for CH 3 asymmetric and CH 3 symmetric were identified at 1453.1 cm −1 and 1381.7 cm −1 , respectively [32,33]. PVA shows the typical infrared bands of hydrocyl groups at 3200-3600 cm −1 which can not be seen in the core-shell structured nanofibers. This indicates that the PVA is not on the surface of the fibers, which is an additional proof of effective encapsulation in core-shell nanofibers [34]. After the PBS treatment, typical infrared bands of hydrocyl groups of PVA at 3200-3600 cm −1 were observed in the nanofibers and FTIR spectra of MH nanofiber are shown in Figure 4a as an example. The SEM image of MH nanofiber is also seen in Figure 4b, and a morphological difference is clearly observed before and after PBS treatment. These observations indicate that distruption of the PLA structure on the surface of the nanofibers provide the presence of PVA on the core.

MTT Assay
As a potential wound dressing material, composite nanofibers should promote cell growth and cell differentiation. The in vitro cytotoxicity tests of coaxial nanofibers toward HaCat cells were evaluated using MTT assays. As shown in Figure 5, the cell viabilities for 72 h were over 80% at all test concentrations, revealing the non-toxicities and good biocompatibilities of composite nanofibers toward cells. Interestingly, we found that neat PVA/PLA nanofibers did not have any significant effect on relative cell viability at 24 h and 48 h time points while in combination with different concentrations of FSP, especially 1% (MHFSP1), it slightly increased the viability, in particularly at 72 h. The cell viabilities of different concentrations had no obvious regularity. After 24 h, the cells may not be fully adherent or have not entered the adaptation process. After 48 h, as the cells enter the adaptation process, some of them may be attached to the material. Some of them may be damaged and dead without attached to the material. Therefore, at the end of 24 h, materials with high viability then may have a low viability after 48 h. However, at the end of 72 h, all test concentrations were found to be above 80% because the cells that were adhering to the material were proliferated. The results suggest that the composite nanofibers did not merely have good biocompatibility but also a promoting effect on cell proliferation after 72 h incubation, indicating that the MH and MHFSP1 composite nanofibers were good biomaterials.

Drug Release Studies and Kinetic Investigation
The amount of MH released from the fibers was determined by UV spectroscopy using a pre-determined calibration curve, y = 0.1178x−0.0083 (R = 0.9898) where y is the MH concentration (in micrograms per milliliter) and x is the solution absorbance at 233 nm which is MH UV absorbance peak. Figure 6 shows the cumulative release rate profiles of MH from coaxial electrospun nanofibers with various FSP concentrations. Experiments of each sample were performed in triplicate, and average values were presented with the error bars, which indicate standard deviation. It is observed from Figure 6 that MH followed a typical dual-stage release profile from the all nanofibers: Firstly, hydrophilic MH showed an initially faster diffusion of coaxial nanofibers (a burst release) and then, a slow trend of sustained release. In fact, it is expected that drug-loaded nanofibers fabricated by coaxial electrospinning generally show sustained release with less initial burst than the blend electrospinning, because the drug is restricted to the core polymer in the core-shell nanofiber while the drug is distributed on the single nanofiber surface. However, faster diffusion or burst release can be observed for the release of hydrophilic small molecule drugs such as MH, as reported previously [35,36].
In the first 24 h, about 75% of the MH was released from the MH nanofibers, while only 59%, 54% and 46% from the MHFSP1, MHFSP3 and MHFSP5 nanofibers, respectively. Furthermore, the nanofibers showed slow and sustained release thereafter and after 21 days where about 99% drug release from MH nanofibers, 96% drug release from MHFSP1 and MHFSP3 nanofibers and 84% drug release from MHFSP3 nanofibers, as shown in Figure 6. These release profiles are in agreement with previous studies of hydrophilic drug release from coaxial nanofibers [37,38]. The results reveal that FSP concentration has a strong effect on initially faster diffusion and sustained drug release. A possible explanation is that the release profile of macromolecular structure of FSP may have formed a suspended site for release of the drug. This improvement can be also attributed to the chemical interactions between MH and FSP content in the nanofiber network. A previous study reported that the increased proportion of the FSP led to increased chemical interactions and thus, a prolonged release behaviour [15]. Clearly as drug carriers, protein loaded nanofiber mats was more stable than protein-free nanofibers. The release kinetics of MH drug was investigated by fitting with various mathematical models such as zero order, first order, Higuchi and Korsmeyer-Peppas models. The obtained kinetic constants and regression coefficients (R 2 ) of all samples are given in Table 2. The first order kinetic have better agreement with higher R 2 values for all samples than other kinetics. Furthermore, in the Korsmeyer-Peppas model, the value of n describes the mechanism of drug release from a polymeric system. For a cylindrical system, 0.45 ≥ n corresponds to a Fickian diffusion and 0.45 < n < 1 refers the Non-Fickian diffusion [39]. As shown in Table 1, all the n values of nanofiber samples are less than 0.45, suggesting MH release from nanofiber samples follow the typical Fickian diffusion mechanism.

Thermal Behaviour
Differential Scanning Calorimetry (DSC) has been widely used to investigate the crystal structures of electrospun nanofiber mats. It is important to understand the possible structural changes of the biomaterial, such as the change in melting temperature, degree of crystallinity, and enthalpy. The DSC thermograms of the Neat, MH, MHFSP1, MHFSP3 and MHFSP5 coaxial nanofibers are presented in Figure 7. The endothermic peak observed at about 150-165 • C for all the samples was attributed to the crystalline melting (T m ) of the nanofibers. Neat-and drug-loaded nanofibers showed double peaks at the melting point, the first one at about 163 • C and the second one at 165 • C. According to previous studies, it seems possible that these results are due to the formation of imperfect crystals during the electrospinning process [40]. DSC thermograms of MHFSP1, MHFSP3 and MHFSP5 samples indicate sharp endothermic peaks at about 150 • C. Table 3 shows the melting point and enthalpy of fusion of coaxial nanofibers for comparison. It can be clearly seen that the presence of FSP caused a decrease in melting point, which indicates that it had a reducing effect on the crystallinity of the PLA-PVA coaxial nanofibers. It was also observed that the melting enthalpy of the nanofibers decreased from 23.5 j/g to 10.8 j/g with the addition of FSP. This result supports that the crystallinity of the nanofibers reduced upon the incorporation of FSP proportion. This result is in agreement with a previous study conducted with a different protein [41] and also confirms our previous findings [24].

Conclusions
In this study, drug and protein loaded nanofibers were successfully fabricated by coaxial electrospinning. Fish Sarcoplasmic Protein (FSP) was produced from Atlantic Bonito and incorported into core polymeric matric at different concentrations. The morphology and chemical compositions of the nanofiber samples were observed using a Scanning Electron Microscopy (SEM) and Fourier Transform Infrared Spectrometer (FTIR). Furthermore, Differential scanning calorimetry (DSC) studies were carried out to investigate the thermal characteristics of the nanofiber mats. According to the results, the coaxial nanofiber mats showed uniform bead-free fiber morphology at an average 650 nm and presented characteristic peaks of the content in the FTIR study. DSC results indicate that the crystallinity of the coaxial nanofibers was reduced with the presence of FSP. The coaxial nanofibers also showed suitable properties for proliferation, and attachment of HaCaT cell lines. Drug release from the coaxial nanofibers were followed by UV spectroscopy and the release profiles were fitted to four popular mathematical equationsm such as zero order, first order, Higuchi and Peppas-Korsemeyer equations to understand the kinetics of drug release. The first order kinetic have better agreement with higher R 2 values for all the samples than other kinetics. Indeed, release kinetics of the nanofibers were described on the typical Fickian diffusion mechanism according to the Peppas-Korsemeyer kinetic. Overall, this study reveals that the composite nanofibers had a good potential to be used in the treatment of diabetic ulcers.