Natural Antibacterial Reagents (Centella, Propolis, and Hinokitiol) Loaded into Poly[(R)-3-hydroxybutyrate-co-(R)-3-hydroxyhexanoate] Composite Nanofibers for Biomedical Applications

Centella asiatica, propolis, and hinokitiol, as natural antibacterial reagents, were integrated into the poly[(R)-3-hydroxybutyrate-co-(R)-3-hydroxyhexanoate] (PHBH) polymer to produce antibacterial wound dressings, using electrospinning process. The results showed that the fiber diameters and surface morphology of PHBH composite nanofibers were influenced by the addition of ethanol–centella (EC), methanol–centella (MC), ethanol–propolis (EP), and ethanol–hinokitiol (EH) at various ratios compared to pristine PHBH nanofibers. From FT-IR, the nanofibrous samples with higher contents of natural antibacterial substances showed the peaks of carboxylic acid, aromatic ring, and tropolone carbon ring from centella, propolis, and hinokitiol, respectively. Furthermore, the tensile strength of neat PHBH nanofibers was increased from 8.00 ± 0.71 MPa up to 16.35 ± 1.78 MPa by loading of propolis (EP) 7% into PHBH. X-ray analysis explained that the loading of propolis (EP) was also able to increase the crystallinity in PHBH composite nanofibers from 47.0% to 54.5%. The antibacterial results demonstrated that PHBH composite nanofibers containing natural antibacterial products were potent inhibitors against the growth of Escherichia coli and Staphylococcus aureus, amongst them hinokitiol and propolis proved to be the most effective. Additionally, the release studies displayed that centella and hinokitiol had faster release from PHBH composite nanofibers in comparison to propolis.


Introduction
Electrospinning is a straightforward and efficient method to fabricate nonwoven mats with continuous fibers in the range from micrometer down to nanometer. Nowadays, nanofibers produced by electrospinning have received the considerably high attention to be an excellent candidate for many important applications in medical field such as drug delivery [1], oral cavity [2], tissue engineering [3-6], wound dressing and healing [7,8] due to the ultra-fine diameter, high surface area to volume ratio, and cost-effectiveness. A large number of studies devoted to utilizing nanofibers that could mimic the extracellular matrix (ECM) of the body and cellular activity [9], as well as sustained drug action such as in skin regeneration or healing [1,10,11] have been published over the last few years. Open wounds were reported to be healed in a much faster-recovering pace by covering from infection using antimicrobials. Without covering, the healing process could be hindered by bacterial interference with cell-matrix interactions, followed by the delay of the cell proliferation and tissue regeneration [12].
The concentrations of natural products in ethanol-centella (15EC and 30EC), methanol-centella (15MC and 30MC), acetone-propolis (10AP), and ethanol-propolis (10EP and 30EP) were determined from the dry weight of each natural product by evaporating solutions under vacuum. Ethanol-hinokitiol (30EH) mentioned the same as 30% because of no insoluble substances in the ethanol solution.
These percentage concentrations of centella, propolis, hinokitiol are displayed in Table 1. These solutions were added to PHBH HFIP solutions with different ratios before the electrospinning process (The ratios will be explained in the next step of preparation of PHBH with natural product solutions). Table 1. The concentrations of natural products in each solution after dry vacuum and ratios of natural product solutions to total weight of spinning poly[(R)-3-hydroxybutyrate-co-(R)-3-hydroxyhexanoate] (PHBH) solutions.

Preparation of PHBH with Natural Antibacterial Reagent Solutions for Electrospinning
PHBH was dissolved in HFIP at room temperature for 4 h to make solution 2 wt%. The natural antibacterial solutions were added into PHBH HFIP solutions and stirred continuously for 16-20 h to get homogeneous solutions. The ratios of ethanol-centella (15EC and 30EC) or methanol-centella (15MC and 30MC) were 1%, 5%, and 10% over the total weight of spinning PHBH solutions. Furthermore, ratios of acetone-propolis (10AP), ethanol-propolis (10EP and 30EP), and ethanol-hinokitiol (EH) to total weight of spinning PHBH solutions were 1%, 5%, and 7%, Table 1. These prepared blend solutions were contained into a glass syringe (10 mL) with 0.6 mm diameter of stainless steel needle. The electrospinning process was carried out using a high voltage power supply by Kato Tech, Japan. The solution feeding rate was 0.1-0.3 mm/min, 15-20 kV voltage, and 15 cm tip-to-collector distance. The environmental conditions of the electrospinning chamber were at a humidity level of 20-30 RH% and temperature ranged between 23-26 • C. The electrospun fibers were collected on the surface of the collector followed by drying at room temperature for 24 h to evaporate the solvents.

Scanning Electron Microscopy (SEM)
The morphology of electrospun nanofibers was investigated by scanning electron microscopy (SEM, JSM-6010LA, JOEL, Tokyo, Japan) at an accelerating voltage of 10 kV with various magnifications. The samples were coated with platinum (Pt) in a sputtering device for 60 s at 30 mA. Image J software was used to measure 50 fibers diameters from each SEM photograph.

Fourier Transform Infrared (FT-IR) Spectroscopy
The presence of centella, propolis, and hinokitiol incorporated in electrospun nanofibers were analyzed using attenuated total reflectance (ATR) FT-IR spectroscopy (IR Prestige-21, Shimadzu Corp., Kyoto, Japan). The infrared spectra were recorded from 4000-600 cm −1 , with a resolution of 4 cm −1 . The infrared beam enters the ATR crystal at an angle of typically 45 • and is totally reflected at the crystal to sample interface with a low signal of noise ratio.

X-ray Diffraction
Wide-angle X-ray diffraction (WAXD) was used for crystalline structure analysis of PHBH nanofibers with or without centella, propolis, or hinokitiol. The two-dimensional (2D) patterns of nanofibers were recorded by X-ray diffraction equipment (SPring-8 synchrotron radiation facility in Japan) with a wavelength of 0.07085 nm at a 2θ scanning angle between 5-30 • . The distance from the sample to the detector (PILATUS 3 × 2 M) was 326.5 mm and the exposure time was 2.0 s. The X-ray diffraction of composite nanofibers was investigated by the MDIP application.

Mechanical Properties
The mechanical properties of neat PHBH, PHBH/centella, PHBH/propolis, PHBH/hinokitiol nanofibers were investigated by using a tensile testing machine (EZTest/EZ-S, Shimadzu Corp., Kyoto, Japan) with 50-N load cell at room temperature. Ten specimens were prepared from each composite nanofiber sheet. The specimens with thickness ranging from 0.05 to 0.14 mm were cut into a rectangular shape of 30 mm × 5 mm. The ends of each sample were secured by two clamps with an initial distance of 10 mm and a crosshead speed of 10 mm/min. Tensile strength, Young's modulus, and elongation at break were calculated on the basis of the stress-strain curves.

Sustained Release Behavior of Natural Antibacterial Reagent from PHBH Composite Nanofibers
The release of centella, propolis, and hinokitiol was analyzed by a UV-visible spectrophotometer (UV2550, Shimadzu Corp., Kyoto, Japan). For release experiments, PHBH composite nanofibers with centella, propolis, or hinokitiol were cut into 2 × 2 cm and measured specimen weight. For each sample, 2 specimens were placed in a test tube filled with 50 mL of phosphate-buffered saline (PBS) (pH 7.4) incubated at 37 • C and stirred at 150 rpm. During the release, 4 mL of the supernatant was retrieved from the release medium at the time points of 5, 10, 20, 30, 40, 50, 60, 120, 240, 360, 480, 600, 720, 1440, 2880, 4320, and 5760 min and the same amount of fresh PBS was added immediately to maintain the medium volume.
The solutions in various concentrations were used to make a standard calibration curve at wavelength 200-700 nm. In detail, certain weight of extract centella, extract propolis, hinokitiol (0.8 mg) was dissolved in 100 mL ethanol/PBS or methanol/PBS (1:99, v/v) and then diluted to 400, 200, 100, 50, 25, 12.5, 6.25, 3.125, and 1.56 µg/mL. UV-vis spectrophotometer at a wavelength of 270 nm, 273 nm, 224 nm was used to measure the amount of released centella, propolis, and hinokitiol from PHBH composite nanofibers, respectively. The total amounts of centella, propolis, and hinokitiol from the nanofiber mats were determined as the average value of the three tests. These results were presented in the form of a cumulative amount of release [30]: where M t is the amount of centella, propolis, or hinokitiol released at the time t, M ∞ is the total amount of centella, propolis, and hinokitiol loaded in PHBH composite nanofibers.

Antibacterial Activity Test
To examine the antibacterial activity of natural antibacterial reagents, S. aureus and E. coli were chosen as representatives for gram-positive and gram-negative bacteria. The test method was the disk diffusion test, which has been well established by previous reports [2,20,[30][31][32][33]. The bacterial culture was spread on the Luria Bertani (LB) agar surface by using a sterile cotton bat. The PHBH/centella, PHBH/propolis, and PHBH/hinokitiol composite nanofibers were placed on the surface of the Petri dishes then they were incubated at 30 • C (E. coli) and 37 • C (S. aureus) for 24 h. The diameters of inhibition zone (mm) were measured for 3 specimens each sample and the results were expressed as mean diameters with standard deviations (in millimeters).

Statistical Analysis
All experiments were conducted in triplicate and the data are presented as mean ± standard deviation (SD). The significant differences were statistically analyzed by one-way analysis of variance (ANOVA) using R free software. Statistical significance was set at p < 0.05 to identify which groups were significantly different from other groups.

Morphology of PHBH Composite Nanofibers with Natural Antibacterial Reagents
The surface morphology of PHBH nanofibers containing different natural antibacterial products was investigated using SEM. Representative SEM images and the fiber diameter distributions of neat PHBH and PHBH nanofibers loaded with different amounts of centella solutions (1%, 5%, and 10%), propolis solutions (1%, 5%, and 7%), and hinokitiol solutions (1%, 5%, and 7%) are exhibited in Figure 1. In general, the different natural product solutions had effects on the surface morphology of composite nanofibers. Table 2 shows the percentage of each natural product solution used, the conditions of the electrospinning process, the average diameters, standard deviations (SD), maximum, and minimum fiber diameters of each sample. From those data, it can be confirmed that PHBH/ethanol-centella (15EC and 30EC) and PHBH/methanol-centella (15MC and 30MC) composite nanofibers showed the uniform fiber diameter. These also exhibited the same trend of a decrease in fiber diameter by increasing the ratio of centella solutions (EC and MC) in spinning PHBH solution. The fiber diameter, with high concentrations of centella solution in PHBH/30EC (10%) and PHBH/30MC (10%), diminished to 349 ± 108 nm and 332 ± 62 nm, respectively. This phenomenon may be due to the decrease in the viscosity of polymer solutions by the addition of ethanol or methanol into HFIP. The viscosity of the mixture solution reduced when the considerable amounts of centella solution added into the polymer matrix [5,14]. Interestingly, the high concentration of centella in PHBH/30EC (10%) ( Figure 1) resulted in fibers merging.
Such a bonded structure was also appeared in PHBH/30EP (7%). The bonded structure in high concentration solutions of EC and EP might be due to using ethanol as the solvent and mixing with HFIP. By contrast, PHBH/30EP (7%) was accompanied by an increase in fiber diameter by loading a high concentration of 30EP. The fiber diameter was increased from 539 ± 99 nm PHBH/30EP (1%) to 739 ± 197 nm PHBH/30EP (7%). A similar behavior was obtained from cellulose acetate nanofibers with a high concentration of honey bee propolis prepared by ethanol solution [33]. Kim et al. [20] also reported that the propolis concentration increased the fiber diameter and provided the linking of PU fibers due to its adhesive properties and demonstrated the optimum utilization of the bonding element for reinforcement of nonwoven fabric. However, different results of fiber diameter were obtained from PHBH composite nanofibers containing acetone-propolis (PHBH/10AP), which are 520 ± 83 nm, 527 ± 161 nm, and 529 ± 109 nm in the cases of 1%, 5%, and 10% (v/v) of 10AP, respectively. This is probably because the evaporation of acetone was faster than ethanol during the electrospinning process.  For PHBH/ethanol-hinokitiol (EH) composite nanofibers, the surface morphology of PHBH/30EH (1%) was smooth, homogeneous without beads. The average diameter was 562 ± 87 nm with a majority of fibers in the range of 438 nm to 745 nm. However, it was difficult to obtain uniform nanofibers with high ratios of hinokitiol solutions (EH), more than 1%. As shown in Figure 1 of PHBH/30EH (5%) and (7%), these resulted in aligned fibers with small grains like beads.
The final surface morphologies of PHBH composite nanofibers were influenced by the kind of used natural products and concentration ratios of natural product solutions to PHBH solutions. This result clearly indicates that the fiber diameters and their distributions can be controlled by concentration ratios of natural products in solutions. In this paper, PHBH composite nanofibers with natural product solutions of PHBH/30EC (10%), PHBH/30MC (10%), PHBH/30EP (7%) and PHBH/30EH (1%) were chosen for further investigation by FTIR, WAXD, mechanical properties, antibacterial test, and release characteristic because these samples had fiber diameter between 300-1000 nm, surface morphology without beads, and high concentration of natural products.

FT-IR Spectral Analysis
The functional groups of natural antibacterial products contained in PHBH composite nanofibers were analyzed by using ATR (FT-IR). The FT-IR spectra of PHBH/30EC (10%), PHBH/30MC (10%), PHBH/30EP (7%), and PHBH/30EH (1%) are shown in Figure 2. The FT-IR spectra of neat PHBH nanofiber and PHBH composite nanofibers with natural products showed similar peaks. The characteristic peaks of PHBH at 2970, 2937, 2870, and 1719 cm −1 were attributed to the stretching vibrations of C-CH 3 , CH 2 , CH, and C=O, respectively [4]. The WAXD measurement was performed in order to confirm the influence of loading centella, propolis, and hinokitiol on the crystallinity of PHBH. All reflections of WAXD and intensity profile are displayed in Figure 3. The neat PHBH nanofibers exhibit diffraction peaks at 2θ = 6.1°, 7.7°, 9°, 10°, 11.7°, 12.4°, and 13.8° which were assigned to the (020), (110), (101), (121), (040), and (002) of the orthorhombic unit cell of PHB crystal, respectively [38]. The intensity of diffraction peaks of neat PHBH nanofibers slightly increased by loading centella and propolis in PHBH/30EC (10%) and PHBH/30EP (7%) composite nanofibers. Moreover, the calculated crystallinity of neat PHBH (47.0%) was slightly lower than that of PHBH/30EC (55.0%), PHBH/30MC (49.2%), and PHBH/30EP (54.5%) in Table 3. It was previously reported that the crystallinity of PHB nanofibers increased by adding the natural phenolic compound caffeic acid [30]. Kim et al. [39] supposed that intermolecular FT-IR graphs of centella powder from ethanol and methanol solutions after filtering and drying exhibit some bands centered at 3320 or 3300 cm −1 corresponding to O-H stretching vibration of carboxylic acid, whereas, 1690 or 1638 cm −1 peaks characterizing C-O stretching vibration. The bands at 1459, 1380, and 1025 cm −1 shows the presence of alkenes with C-H in-plane bending, the stretching vibration of C-N for aromatic amide group, and C-O stretching, respectively (Figure 2A). These spectra of those compounds were confirmed in the research of centella asiatica by Manotham et al. [5], Sondari et al. [34], and Sugunabai et al. [35]. As showed in Figure 2A, it proved that using methanol as solvent yielded higher content of Asiatic acid than using ethanol with the same concentration from raw centella (30%EC and 30%MC). In the case of PHBH/30MC (10%) composite nanofibers, the decline of the band at 3340 cm −1 suggested the formation of the intermolecular hydrogen bond between centella and PHBH polymer.
The FT-IR spectra of propolis powder and PHBH/30EP (7%) nanofibers were exhibited in Figure 2B. In the case of propolis powder, the band at 3340 cm −1 was assigned to stretching vibration of the O-H group in the phenolic compound and the band at 2920-2870 cm −1 was attributed to C-H aliphatic stretching vibration (stretching vibration of CH 2 and CH 3 ) [36]. The bands at 1706 cm −1 , 1650-1602 cm −1 , and 1190 cm −1 were attributed to the C=O group, aromatic ring deformations C=C stretching vibration, and C-O stretching vibration, respectively. In PHBH/30EP (7%) composite nanofibers, some peaks appeared at between 1651 and 1610 cm −1 , and 1510 cm −1 , confirming the presence of propolis in PHBH nanofiber mats. These bands were assigned to stretching of C=C aromatic ring bands in flavonoids [17,36].
The characteristic absorption peaks of hinokitiol powder were observed at 3200, 1609, 1543, 1476, 1459, 1417, 1269, 1185, and 950 cm −1 . These bands represented O-H stretching, C=C and C=O stretching, C=C stretching (in-phase), ring CH bending, C-O-H group, and ring CH bending, respectively [37]. Dyrskov et al. [23] reported that the absorbance bands in the hinokitol spectrum ascribed to C=O stretching in the tropolone carbon ring (1609 and 1543 cm −1 ) and the C-O-H group (1189 cm −1 ). The presence of hinokitiol in PHBH composite nanofibers was confirmed by the weak peak appeared at 1610 cm −1 that is related to tropolone carbon ring. In general, the peak intensity decreased by loading natural product might reveal that natural product is homogeneously distributed through the polymer matrix [20]. The nanofibrous samples with higher contents of natural antibacterial substances showed the peaks of carboxylic acid, aromatic ring, and tropolone carbon ring from centella, propolis, and hinokitiol, respectively.
interaction through a hydrogen bonding of plant polyphenol with polycaprolactone probably increased by slight increment of crystallinity in polycaprolactone. Whereas, the crystallinity of PHBH/30EH (1%) composite nanofibers was declined to 44.1% compared to neat PHBH. These results indicate that an intermolecular interaction between PHBH polymer and centella or propolis in nanofibers mat exists.

Mechanical Characteristic of Composite Nanofibers
Tensile test was conducted on neat PHBH and PHBH composite nanofibers with the natural antibacterial products in order to evaluate the influences of natural product loading. The tensile strength, elongation at break, and Young's modulus of neat PHBH, PHBH/30EC (10%), PHBH/30MC (10%), PHBH/30EP (7%), and PHBH/30EH (1%) are shown in Table 3. Figure 4 illustrates the representative stress-strain curves of PHBH composite nanofiber samples. The tensile strength of neat PHBH was 8.0 MPa with high elongation up to 61.5%. Interestingly, the loading of centella (EC 10%) and propolis (EP 7%) in PHBH demonstrated the same trend with the tensile strength rose up to 17.8 and 16.4 MPa but it reduced elongation down to 15.6% and 9.3%, respectively. PHBH/30MC (10%) composite nanofibers exhibited no significant differences in tensile strength and Young's modulus compared to neat PHBH. PHBH/30EH (1%) resulted in a decline in both tensile strength and elongation compared to neat PHBH.

In Vitro Antibacterial Activity
The antibacterial activity of neat PHBH nanofibers, PHBH composite nanofibers with centella, propolis, and hinokitol against the gram-positive bacteria S. aureus and gram-negative bacteria E. coli was evaluated using inhibition zone method [20,33,40,41]. The antimicrobial effect of samples was evaluated at 37 °C (S. aureus) and 30 °C (E. coli) for 24 h. Table 4 presents the diameters of the inhibition zone for PHBH composite nanofibers with natural products. The effects of natural antibacterial products via the agar diffusion test are clearly shown in Figure 5. PHBH/30EC (10%) was observed to have no effect against gram-positive and gram-negative bacteria. Whereas, centella in PHBH/30MC (10%) demonstrated low antibacterial activity against S. aureus with inhibitory zone of 7.7 mm and no effect against E. coli. The diameter of the inhibition zone of PHBH/30EP (7%) was 18.3 mm for S. aureus and 17.3 mm for E. coli. PHBH/30EH (1%) composite nanofibers have extensive inhibition zone against S. aureus (25.7 mm) and E. coli (29.7 mm), respectively. The zones of inhibition for PHBH/30EP (7%) against S. aureus were slightly broader than that against E. coli. Zeighampour et al. [21] explained that hydroalchoholic extract of propolis showed better antibacterial activities against gram-negative bacteria than gram-positive bacteria due to the different cell wall chemical structures of bacteria. However, the opposite results for PHBH/30EH (1%) might be due to the different modes of action and the bactericidal effects of natural products.
Regarding these results, the antimicrobial effects of PHBH composite nanofibers with propolis and hinokitiol were observed to be better than these of centella. However, the low effect antibacterial activity of centella (EC or MC) might be due to the low chemical content in PHBH composite nanofiber as analysed by FT-IR, which PHBH/30EC (10%) showed no clear detection of asiatic acid. For further PHBH composite nanofibers with centella (EC) and propolis (EP) using ethanol as solvent displayed higher tensile strength in comparison to neat PHBH, PHBH/MC, and PHBH/EH. These phenomena may be due to the surface morphology and polymer structure in PHBH composite nanofibers. The surface morphology of PHBH/30EC (10%) and PHBH/30EP (7%) showed bonded-structure between fibers, whereas PHBH/30MC (10%) and PHBH/30EH (1%) surface revealed randomly oriented fibers. Kim et al. [20] reported the similar behaviors of merged structure, the adhesive properties of propolis might be useful to increase the mechanical strength of polyurethane fibers. The mechanical properties of nonwoven nanofibers also depended on parameters such as surface interaction among the fibers, average fiber diameter, fiber defects (bead formation) during electrospinning [10,11].
Additionally, the enhancement of physical properties in PHBH/30EC (10%) and PHBH/30EP (7%) might be related to the rise of the crystallinity of PHBH polymer in PHBH composite nanofibers. It is probably due to the interaction of natural products (EC and EP) to PHBH polymer that was explained by FT-IR and X-ray analysis.

In Vitro Antibacterial Activity
The antibacterial activity of neat PHBH nanofibers, PHBH composite nanofibers with centella, propolis, and hinokitol against the gram-positive bacteria S. aureus and gram-negative bacteria E. coli was evaluated using inhibition zone method [20,33,40,41]. The antimicrobial effect of samples was evaluated at 37 • C (S. aureus) and 30 • C (E. coli) for 24 h. Table 4 presents the diameters of the inhibition zone for PHBH composite nanofibers with natural products. The effects of natural antibacterial products via the agar diffusion test are clearly shown in Figure 5. PHBH/30EC (10%) was observed to have no effect against gram-positive and gram-negative bacteria. Whereas, centella in PHBH/30MC (10%) demonstrated low antibacterial activity against S. aureus with inhibitory zone of 7.7 mm and no effect against E. coli. The diameter of the inhibition zone of PHBH/30EP (7%) was 18.3 mm for S. aureus and 17.3 mm for E. coli. PHBH/30EH (1%) composite nanofibers have extensive inhibition zone against S. aureus (25.7 mm) and E. coli (29.7 mm), respectively. The zones of inhibition for PHBH/30EP (7%) against S. aureus were slightly broader than that against E. coli. Zeighampour et al. [21] explained that hydroalchoholic extract of propolis showed better antibacterial activities against gram-negative bacteria than gram-positive bacteria due to the different cell wall chemical structures of bacteria. However, the opposite results for PHBH/30EH (1%) might be due to the different modes of action and the bactericidal effects of natural products.    Regarding these results, the antimicrobial effects of PHBH composite nanofibers with propolis and hinokitiol were observed to be better than these of centella. However, the low effect antibacterial activity of centella (EC or MC) might be due to the low chemical content in PHBH composite nanofiber as analysed by FT-IR, which PHBH/30EC (10%) showed no clear detection of asiatic acid. For further work, raising the centella concentrations in ethanol or methanol solutions might be one of the solutions to increase the antibacterial activity, if those will be used for wound healing application. Overall, the natural products (especially propolis and hinokitiol) loaded into PHBH composite nanofibers inhibited the growth of S. aureus and E. coli powerfully.

Release Behavior of Natural Antibacterial Product
The release profiles of centella, propolis, and hinokitiol from electrospun PHBH composite nanofibers were plotted as a function of time in PBS with pH 7.4 at 37 • C. The release curves of PHBH/30EC (10%), PHBH/30MC (10%), PHBH/30EP (7%), and PHBH/EH (1%) composite nanofibers demonstrated different release behaviors in each sample as shown in Figure 6. Centella from PHBH/30EC started to release within the 10 min and completely released in 20 min. In the case of PHBH/30MC, Figure 6B, the similar behaviors were also observed with complete release in 20 min. The maximum releases of centella from PHBH/30EC and PHBH/30MC were 13.9 % and 31.5%, respectively. While the significant release amounts of propolis from PHBH/30EP started to be detected after 10 min and gradually increased up to 2880 min (48 h) with the release amount of 9.6%. The release of propolis from PHBH/30EP sustained longer than others but the release amount of propolis was less. The fast release of hinokitiol from PHBH/30EH was noticeable in the first 5 min and gradually increased up to 20 min with a maximum release of 46.1%. Afterwards, the release of hinokitiol gradually declined over 240 min (4 h). This decline might be due to the degradation of hinokitiol in PBS solution by light, heat, or solvent (ethanol). In the study of ciprofloxacin hydrochloride (CpHCl) release from electrospun alginate [42], at around 24% release of total loaded drug, the cross-linking process between CpHCl molecules in the nanofibers happened and affected the release behaviors. This result also suggests the relation between the release capacity and the characteristics of PHBH, which is hydrophobic polymer. Ignatova et al. [43] reported that the release of caffeic acid phenethyl ester (CAPE) in material was influenced by hydrophilic-hydrophobic features of the fibrous mats, the release of CAPE increased markedly when CAPE was incorporated in PVP (hydrophilic) matrix than when incorporated in PHB (hydrophobic) matrix. Especially for PHBH/30EC (10%) and PHBH/30EP (7%), those have similar release behaviors, when only small amounts of natural products were released to PBS. It can be considered due to the high crystallinity of those PHBH composite nanofibers that affected release characteristics. In comparison, PHBH/30EH (1%) with low crystallinity showed fast and high release amounts. Study of drug release of ampicillin incorporated poly(methyl methacrylate)-nylon6 core/shell nanofibers proved that the increase of released drug, most probably due to decreased crystallinity in the polymer matrix [44].

Conclusions
The natural antibacterial products (centella, propolis, and hinokitiol) loaded into PHBH composite nanofibers were successfully fabricated by the electrospinning process. The fiber diameter and surface morphology of PHBH composite nanofibers with antibacterial reagents can be controlled by different kinds of natural antibacterial products, solvent systems for dissolution, and component concentrations in the PHBH-HFIP polymer solutions. The presence of carboxylic acid and aromatic amide groups in PHBH/centella, aromatic ring bands due to flavonoid in PHBH/propolis, and tropolone carbon ring in PHBH/hinokitiol were confirmed by FT-IR. The loading of centella and propolis led to an increase in the crystallinity of the PHBH polymer. Furthermore, the loading of centella and propolis improved the tensile strength, compared to neat PHBH nanofibers. Hinokitiol and propolis were proved to be potent antibiotics by large inhibition zones against both E. coli and S. aureus. The release of centella and hinokitiol from PHBH nanofibers was fast and finished in 20 min with maximum release of 13.9% (PHBH/30EC), 31% (PHBH/30MC), and 46% (PHBH/30EH). Whereas, the release of propolis was continuously to 48 h and maximal release was 9.5%. These results in our study indicated that natural antibacterial products loaded into PHBH composite nanofibers can improve the mechanical properties (PHBH/30EP and PHBH/30EC) and prove the composite nanofibers antibacterial effects against the gram-negative and gram-positive bacteria (PHBH/30EP and PHBH/30EH), which are important characteristics as wound healing materials. Interestingly, PHBH/30EC (10%) showcased good mechanical properties and PHBH/30MC (10%) exhibited good morphology in nanofiber form, however, it may necessary to increase the concentration of centella in centella solution (ethanol or methanol) if it is to be used as antibacterial reagents. PHBH/30EH (1%) might be used as drug delivery with rapid release but possesses low mechanical properties. PHBH/30EP (7%) could be used in wound healing with needed mechanical properties and slow release for long-time effects.