Antibacterial Electrospun Polycaprolactone Nanofibers Reinforced by Halloysite Nanotubes for Tissue Engineering

Due to its slow degradation rate, polycaprolactone (PCL) is frequently used in biomedical applications. This study deals with the development of antibacterial nanofibers based on PCL and halloysite nanotubes (HNTs). Thanks to a combination with HNTs, the prepared nanofibers can be used as low-cost nanocontainers for the encapsulation of a wide variety of substances, including drugs, enzymes, and DNA. In our work, HNTs were used as a nanocarrier for erythromycin (ERY) as a model antibacterial active compound with a wide range of antibacterial activity. Nanofibers based on PCL and HNT/ERY were prepared by electrospinning. The antibacterial activity was evaluated as a sterile zone of inhibition around the PCL nanofibers containing 7.0 wt.% HNT/ERY. The morphology was observed with SEM and TEM. The efficiency of HNT/ERY loading was evaluated with thermogravimetric analysis. It was found that the nanofibers exhibited outstanding antibacterial properties and inhibited both Gram- (Escherichia coli) and Gram+ (Staphylococcus aureus) bacteria. Moreover, a significant enhancement of mechanical properties was achieved. The potential uses of antibacterial, environmentally friendly, nontoxic, biodegradable PCL/HNT/ERY nanofiber materials are mainly in tissue engineering, wound healing, the prevention of bacterial infections, and other biomedical applications.


Introduction
Halloysite nanotubes (HNTs) are natural, nontoxic, biocompatible, eco-friendly, and low-cost materials recognized by the Environmental Protection Agency as nanomaterials (EPA 4). Presently, HNTs play a significant role in drug-carrier systems suitable for different biomedical applications, e.g., tissue engineering [1,2].
HNTs, naturally occurring in 1:1 layered aluminosilicate clay, consist of aluminum and silicon oxide layers rolled into tubes. The layers are rolled into tubes because of differences in the sizes of silicon and aluminum ions [3]. The typical length of the nanotubes is about 1-2 µm. Their outer and inner diameters range from 50 to 100 nm and from 10 to 50 nm, respectively [4,5]. For biomedical applications, the most significant advantages of HNTs

Loading of a Drug in Halloysite Nanotubes
The loading of active agents to halloysite is based on the diffusion of molecules from an external solution into the inner part of HNTs due to the concentration gradient. The evaporation of the solvent under the vacuum elevates the concentration of the active agents in the solution and enhances the diffusion rate. Therefore, fast-drying solvents with low viscosities such as acetone or ethanol are preferable for organic substances [38]. To eliminate potential water on the outer part of HNTs before the loading procedure, HNTs were dried in the oven for 2 h at 150 • C. Since ERY is soluble in methanol, after the drying, it was dissolved in methanol. HNTs were dispersed in the ERY solution and sonicated for 20 min. The rate of halloysite to erythromycin was 60:40. After the sonification, halloysite loaded with ERY was dried in a vacuum and washed with distilled water using a 0.2 µm membrane filter. Residual unloaded ERY was removed with methanol (4 h).
HNTs were dried in the oven for 2 h at 150 • C. We dissolved 400 mg of the drug in 25 mL of methanol. We dispersed 600 mg of dried halloysite in a drug solution (ratio 40:60 ERY:HNT), which was then sonicated for 20 min. Then, this dispersion was dried in a vacuum. After vacuum drying, it was washed with distilled water using a 0.2 µm membrane filter. Residual unloaded ERY was removed with methanol (4 h).

Optimization of Electrospinning Process Parameters
The electrospinning process for PCL composites (Figure 1) was optimized and previously described by our group [29]. Briefly, a solution of 10% w/v PCL was prepared in a mixture of chloroform and methanol in a ratio of 4:1 and continuously stirred for 2 h. Then, the mixture was kept unstirred for another 15 min to remove any bubbles present in the solution. Prepared HNTs loaded with ERY were added to the PCL solution and stirred for another 1 h. Electrospun fibers were fabricated using a Spellman high voltage power source (Spellman High Voltage Electronics Corporation, New York, NY, USA) and syringe pump (New Era Pump Systems, Inc., New York, NY, USA). The electrospinning apparatus consisted of a 10 mL syringe that was integrated with a grounded electrode, and the needle diameter was 0.41 mm. The distance between tee collector and source was kept to 13 cm. One thin aluminum sheet was fixed over the static collector. The feeding rate for the electrospinning solution was set to 1 mL/h at a voltage of 25 kV. The temperature for the location varied between 21 and 29 • C, and the humidity varied from 73% to 93%.

Scanning Electron Microscopy
For the needs of scanning electron microscopy (SEM), the sample's surface was covered in gold with a Sample Preparation System Quorum Technologies Q150R S/E/ES sputter coater evaporator (Quorum Technologies, Laughton, England, and the micrographs were obtained with an FIB Microscope Quanta 3D 200i (FEI Company, Tokyo, Japan) in a secondary electron mode at different magnitudes. The surface morphologies of pure and modified fibers were obtained.

Transmission Electron Microscopy
The structure of HNTs loaded with ERY was characterized by transmission electron microscopy (Jeol TEM 1200EX, JEOL Ltd., Tokyo, Japan at an accelerating voltage of 100 kV. The sample was dispersed on a copper grid with carbon support film.

Antibacterial Activity
The antibacterial activity of HNT/ERY before and after removing residual ERY from HNTs was assayed. The antibacterial activity of PCL/HNT/ERY nanofibers was assessed with the diffusion method by placing the HNT, HNT:ERY, 1 cm 2 of prepared PCL fiber, PCL/HNT fiber, and PCL/HNT and ERY fiber on the inoculated (10 6 cells/mL) MHA (Mueller Hinton Agar) growth media. For inoculation, the model bacteria S. aureus CCM 3953 (Czech Collection of Microorganisms) and E. coli CCM 3988 were used. The antibacterial activity of prepared nanofibers was evaluated as the occurrence of a sterile zone of inhibition around the modified HNTs with ERY and modified fibers of PCL/HNT and ERY, respectively, after cultivation for 24 h at 37 °C. With the aim to wash out the residual ERY from the surface of the prepared materials, modified HNTs with ERY and the PCL/HNT and ERY fibers were washed two times in methanol and two times in water. The antibacterial activity was evaluated again as described above. To evaluate the sustained release of ERY from the PCL/HNT fiber, serial cultivation was examined as follows. The fresh prepared PCL/HNT and ERY fibers were first placed on the inoculated growth media, and the antibacterial effect was evaluated after 24 h of cultivation at 37 °C. Then, these same fibers from the grown bacterial cultures were replaced with fresh ones that were again cultivated and considered. This procedure was repeated for 5 days.

Mechanical Property Measurement
Mechanical properties were tested in uniaxial tension at a crosshead speed of 5 mm•min −1 and ambient temperature of 22 °C using a Zwick Roell Z010 (Zwick-Roell, Ulm, Germany) equipped with a 10 kN load cell. Six specimens of rectangle shape (approximately 30 × 5 mm 2 ) were tested for each type of material, and the averages and standard

Scanning Electron Microscopy
For the needs of scanning electron microscopy (SEM), the sample's surface was covered in gold with a Sample Preparation System Quorum Technologies Q150R S/E/ES sputter coater evaporator (Quorum Technologies, Laughton, England, and the micrographs were obtained with an FIB Microscope Quanta 3D 200i (FEI Company, Tokyo, Japan) in a secondary electron mode at different magnitudes. The surface morphologies of pure and modified fibers were obtained.

Transmission Electron Microscopy
The structure of HNTs loaded with ERY was characterized by transmission electron microscopy (Jeol TEM 1200EX, JEOL Ltd., Tokyo, Japan) at an accelerating voltage of 100 kV. The sample was dispersed on a copper grid with carbon support film.

Antibacterial Activity
The antibacterial activity of HNT/ERY before and after removing residual ERY from HNTs was assayed. The antibacterial activity of PCL/HNT/ERY nanofibers was assessed with the diffusion method by placing the HNT, HNT:ERY, 1 cm 2 of prepared PCL fiber, PCL/HNT fiber, and PCL/HNT and ERY fiber on the inoculated (10 6 cells/mL) MHA (Mueller Hinton Agar) growth media. For inoculation, the model bacteria S. aureus CCM 3953 (Czech Collection of Microorganisms) and E. coli CCM 3988 were used. The antibacterial activity of prepared nanofibers was evaluated as the occurrence of a sterile zone of inhibition around the modified HNTs with ERY and modified fibers of PCL/HNT and ERY, respectively, after cultivation for 24 h at 37 • C. With the aim to wash out the residual ERY from the surface of the prepared materials, modified HNTs with ERY and the PCL/HNT and ERY fibers were washed two times in methanol and two times in water. The antibacterial activity was evaluated again as described above. To evaluate the sustained release of ERY from the PCL/HNT fiber, serial cultivation was examined as follows. The fresh prepared PCL/HNT and ERY fibers were first placed on the inoculated growth media, and the antibacterial effect was evaluated after 24 h of cultivation at 37 • C. Then, these same fibers from the grown bacterial cultures were replaced with fresh ones that were again cultivated and considered. This procedure was repeated for 5 days.

Mechanical Property Measurement
Mechanical properties were tested in uniaxial tension at a crosshead speed of 5 mm·min −1 and ambient temperature of 22 • C using a Zwick Roell Z010 (Zwick-Roell, Ulm, Germany) equipped with a 10 kN load cell. Six specimens of rectangle shape (approximately 30 × 5 mm 2 ) were tested for each type of material, and the averages and standard deviations were determined. The thickness of the specimens was measured with a micrometer.

Thermogravimetric Analysis
Thermal stability was determined by thermogravimetric analysis (TGA) using a Q1500 D instrument (from MON Budapest, with TA Universal Analysis software). Each sample (100 mg) was heated from 30 to 600 • C at a heating rate of 10 • C/min in the presence of air with a flow of 50 mL/min. The corresponding weight loss was recorded as a function of temperature.

Microscopic HNT/ERY' Structure Analysis
As our task was the preparation of antimicrobial nanofibers for biomedical applications, the main criterion for the selection of an appropriate type of halloysite was its purity. For this reason, we selected ULTRA HalloPureTM tubular halloysite. This particular HNT type has a very high purity (93.5%), with only trace levels of feldspar. The second reason for our study was ULTRA HalloPure's geometry. From the point of view reinforcing effect, which is essential for nanofibers based on low-strength PCL, it is necessary to use HNTs with long aspect ratios. Moreover, long-aspect-ratio HNTs provide more prolonged antibacterial effects due to the release of an active compound. The geometry of the used HNT type (length, inner and outer diameter, and aspect ratio) is presented in Table 1.
The micromorphology and nanomorphology of HNTs both modified and loaded with ERY were studied with transmission electron microscopy. The structure of HNTs before ERY loading is shown in Figure 2. TEM images of HNTs loaded with ERY are shown in Figure 3. deviations were determined. The thickness of the specimens was measured with a micrometer.

Thermogravimetric Analysis
Thermal stability was determined by thermogravimetric analysis (TGA) using a Q1500 D instrument (from MON Budapest, with TA Universal Analysis software). Each sample (100 mg) was heated from 30 to 600 °C at a heating rate of 10 °C/min in the presence of air with a flow of 50 mL/min. The corresponding weight loss was recorded as a function of temperature.

Microscopic HNT/ERY' Structure Analysis
As our task was the preparation of antimicrobial nanofibers for biomedical applications, the main criterion for the selection of an appropriate type of halloysite was its purity. For this reason, we selected ULTRA HalloPureTM tubular halloysite. This particular HNT type has a very high purity (93.5%), with only trace levels of feldspar. The second reason for our study was ULTRA HalloPure's geometry. From the point of view reinforcing effect, which is essential for nanofibers based on low-strength PCL, it is necessary to use HNTs with long aspect ratios. Moreover, long-aspect-ratio HNTs provide more prolonged antibacterial effects due to the release of an active compound. The geometry of the used HNT type (length, inner and outer diameter, and aspect ratio) is presented in Table  1.
The micromorphology and nanomorphology of HNTs both modified and loaded with ERY were studied with transmission electron microscopy. The structure of HNTs before ERY loading is shown in Figure 2. TEM images of HNTs loaded with ERY are shown in Figure 3.   deviations were determined. The thickness of the specimens was measured with a micrometer.

Thermogravimetric Analysis
Thermal stability was determined by thermogravimetric analysis (TGA) using a Q1500 D instrument (from MON Budapest, with TA Universal Analysis software). Each sample (100 mg) was heated from 30 to 600 °C at a heating rate of 10 °C/min in the presence of air with a flow of 50 mL/min. The corresponding weight loss was recorded as a function of temperature.

Microscopic HNT/ERY' Structure Analysis
As our task was the preparation of antimicrobial nanofibers for biomedical applications, the main criterion for the selection of an appropriate type of halloysite was its purity. For this reason, we selected ULTRA HalloPureTM tubular halloysite. This particular HNT type has a very high purity (93.5%), with only trace levels of feldspar. The second reason for our study was ULTRA HalloPure's geometry. From the point of view reinforcing effect, which is essential for nanofibers based on low-strength PCL, it is necessary to use HNTs with long aspect ratios. Moreover, long-aspect-ratio HNTs provide more prolonged antibacterial effects due to the release of an active compound. The geometry of the used HNT type (length, inner and outer diameter, and aspect ratio) is presented in Table  1.
The micromorphology and nanomorphology of HNTs both modified and loaded with ERY were studied with transmission electron microscopy. The structure of HNTs before ERY loading is shown in Figure 2. TEM images of HNTs loaded with ERY are shown in Figure 3.

Loading Efficiency of HNTs
A thermogravimetric study of HNTs and HNT/ERY was performed to determine loading efficiency, i.e., the amount of ERY loaded into HNTs. As mentioned earlier, the loading of active agents to HNTs is based on the diffusion of molecules from external solutions into the inner part of HNTs due to the concentration gradient. As shown in Figure 3, the unmodified HNTs were stable up to 500 • C, at which point de-hydroxylation [6,39] occurred with a disruption of the tube-wall multilayer packing. At 600 • C, the weight loss of the unmodified HNTs was 12.5 wt.% (blue curve in Figure 4). Pure erythromycin showed steep decomposition around 300 • C followed by a slow degradation process up to 600 • C. (red curve in Figure 4). The efficiency of the used loading method was evaluated via a comparison TGA of pristine HNTs with HNTs loaded with ERY (HNT/ERY) washed with water (green curve in Figure 4) and methanol (orange curve in Figure 4).
The sample of HNT/ERY washed in water (HNT/ERY/WAT) showed the main decomposition around 300 • C followed by gradual degradation process and further degradation step around 500 • C, reflecting the behavior of precursors.

Loading Efficiency of HNTs
A thermogravimetric study of HNTs and HNT/ERY was performed to determine loading efficiency, i.e., the amount of ERY loaded into HNTs. As mentioned earlier, the loading of active agents to HNTs is based on the diffusion of molecules from external solutions into the inner part of HNTs due to the concentration gradient. As shown in Figure 3, the unmodified HNTs were stable up to 500 °C, at which point de-hydroxylation [6,39] occurred with a disruption of the tube-wall multilayer packing. At 600 °C, the weight loss of the unmodified HNTs was 12.5 wt.% (blue curve in Figure 4). Pure erythromycin showed steep decomposition around 300 °C followed by a slow degradation process up to 600 °C. (red curve in Figure 4). The efficiency of the used loading method was evaluated via a comparison TGA of pristine HNTs with HNTs loaded with ERY (HNT/ERY) washed with water (green curve in Figure 4) and methanol (orange curve in Figure 4).
The sample of HNT/ERY washed in water (HNT/ERY/WAT) showed the main decomposition around 300 °C followed by gradual degradation process and further degradation step around 500 °C, reflecting the behavior of precursors. The weight loss of HNT/ERY/WAT-37.9 wt.% at 500 °C-corresponded to the rate of erythromycin loaded on both the inner and outside surface of HNTs (HNT:ERY at 60:40) and the rate of ERY loaded exclusively onto the inner surface of HNTs, as evaluated on the sample from which residual ERY was removed with methanol: 4.6 and 11.5 vol.%, respectively. In summary, we obtained extremely highly-loaded HNTs suitable for antibacterial applications in biomedicine.

Morphological Observation of PCL/HNT/ERY Nanofibers
Our previous study described the morphology of electrospun PCL antibacterial composites with different fillers-hydrophobic quantum dots working on the principle of photodynamic therapy-for wound healing in tissue engineering in detail [29]. A strongly porous material with interconnected structures was obtained via the electrospinning method. This 3D structure is very suitable for use in tissue engineering. Despite this fact, we created relatively homogeneous nanofibers with an average thickness of 3-4 μm, and the same effect was not observed in the case of HNT-loaded samples. As shown in Figure  5, the fibers contained a lot of beads in their structure due to the very high probability of  The weight loss of HNT/ERY/WAT-37.9 wt.% at 500 • C-corresponded to the rate of erythromycin loaded on both the inner and outside surface of HNTs (HNT:ERY at 60:40) and the rate of ERY loaded exclusively onto the inner surface of HNTs, as evaluated on the sample from which residual ERY was removed with methanol: 4.6 and 11.5 vol.%, respectively. In summary, we obtained extremely highly-loaded HNTs suitable for antibacterial applications in biomedicine.

Morphological Observation of PCL/HNT/ERY Nanofibers
Our previous study described the morphology of electrospun PCL antibacterial composites with different fillers-hydrophobic quantum dots working on the principle of photodynamic therapy-for wound healing in tissue engineering in detail [29]. A strongly porous material with interconnected structures was obtained via the electrospinning method. This 3D structure is very suitable for use in tissue engineering. Despite this fact, we created relatively homogeneous nanofibers with an average thickness of 3-4 µm, and the same effect was not observed in the case of HNT-loaded samples. As shown in Figure 5, the fibers contained a lot of beads in their structure due to the very high probability of the electric conductivity of the electrospinning solution to change, which is one of the key factors of the electrospinning process. Hydrophobic quantum dots are semiconductors, and HNTs or ERY-loaded HNTs are insulators. Therefore, the final electric conductivity of the solution was significantly lower and the viscosity of solution was reduced. All those Polymers 2022, 14, 746 7 of 12 parameters affected the final structure of the fibers. These beads had only a minor effect on antibacterial and mechanical properties because these were determined by releasing ERY from HNTs, as shown later.
Polymers 2022, 14, x FOR PEER REVIEW 7 of 12 the electric conductivity of the electrospinning solution to change, which is one of the key factors of the electrospinning process. Hydrophobic quantum dots are semiconductors, and HNTs or ERY-loaded HNTs are insulators. Therefore, the final electric conductivity of the solution was significantly lower and the viscosity of solution was reduced. All those parameters affected the final structure of the fibers. These beads had only a minor effect on antibacterial and mechanical properties because these were determined by releasing ERY from HNTs, as shown later.

Antibacterial Properties of PCL/Halloysite/ERY
The antibacterial activity of HNT/ERY before and after the removal of residual ERY from HNT surfaces was assayed (Figure 6), and the obtained antibacterial properties of PCL/HNT and ERY are shown in Figure 7. Because PCL and HNT/PCL were not antibacterial active, as is clear from Figure 7, no inhibition zone was formed; the HNT/ERY and PCL/HNT/ERY nanofibers revealed significant antibacterial activity.

Antibacterial Properties of PCL/Halloysite/ERY
The antibacterial activity of HNT/ERY before and after the removal of residual ERY from HNT surfaces was assayed (Figure 6), and the obtained antibacterial properties of PCL/HNT and ERY are shown in Figure 7. Because PCL and HNT/PCL were not antibacterial active, as is clear from Figure 7, no inhibition zone was formed; the HNT/ERY and PCL/HNT/ERY nanofibers revealed significant antibacterial activity.
Zones of inhibition were observed around the HNT and ERY fibers and the = PCL/HNT and ERY fibers. After the successful incorporation of ERY into the carrier HNTs (proving that it was also a part of the prepared PCL/HNT fibers), the washing of the fibers in methanol followed by washing in water was conducted to ensure that the whole residual ERY of the surface was removed. Next, the antibacterial assay was repeated with washed materials. As is clearly shown in Figures 6 and 7, even though the zones of inhibition of washed samples were smaller, antibacterial activity was recorded again, which means that the antibiotic was incorporated into the HNT nanostructures. Accordingly, HNTs are an appropriate carrier for ERY.
Finally, the antibacterial activity was assayed when the PCL/HNT and ERY fibers were repeatedly (three times) replaced on fresh inoculated growth media. The results are shown in Table 2. These results support the idea of sustained release from the prepared fiber because the antibacterial effect was still observed after the second replacement (72 h) for E. coli and after the first replacement (48 h) for S. aureus (Table 2).  Zones of inhibition were observed around the HNT and ERY fibers and the = PCL/HNT and ERY fibers. After the successful incorporation of ERY into the carrier HNTs (proving that it was also a part of the prepared PCL/HNT fibers), the washing of the fibers in methanol followed by washing in water was conducted to ensure that the whole residual ERY of the surface was removed. Next, the antibacterial assay was repeated with washed materials. As is clearly shown in Figures 6 and 7, even though the zones of inhibition of washed samples were smaller, antibacterial activity was recorded again, which means that the antibiotic was incorporated into the HNT nanostructures. Accordingly, HNTs are an appropriate carrier for ERY.  Zones of inhibition were observed around the HNT and ERY fibers and the = PCL/HNT and ERY fibers. After the successful incorporation of ERY into the carrier HNTs (proving that it was also a part of the prepared PCL/HNT fibers), the washing of the fibers in methanol followed by washing in water was conducted to ensure that the whole residual ERY of the surface was removed. Next, the antibacterial assay was repeated with washed materials. As is clearly shown in Figures 6 and 7, even though the zones of inhibition of washed samples were smaller, antibacterial activity was recorded again, which means that the antibiotic was incorporated into the HNT nanostructures. Accordingly, HNTs are an appropriate carrier for ERY.

Mechanical Properties of PCL/HNT/ERY Nanofibers
As it is clearly evident, the mechanical properties of investigated PCL nanofiber system were significantly influenced by the presence of both unmodified and HNT-loaded ERY ( Figure 8, Table 3). The introduction of HNT particles into nanofibers significantly increased the Young's modulus and tensile strength. Furthermore, elongation at break remained unchanged for this system. The observed enhancement correlated well with the results obtained on PCL [40]. Stress transfer, volume replacement, and segmental immobilization [41][42][43] reinforcing mechanisms were responsible for this enhancement of mechanical performance. The loading of ERY into HNTs decreased Young's modulus to the value of the PCL while elongation at break was significantly increased. The observed trend re-affirmed the effect of plasticizers-small compatible molecules that are able to decrease the glass transition temperature of polymers-and suggested the good compatibility of ERY with the PCL/HNT nanofibers. More importantly, tensile strength also decreased upon ERY loading. However, the decrease was so slight that the values of the PCL/HNT and ERY systems were still more than 100% higher than those of the pure PCL system. A similar tensile strength enhancement (about 100%) was observed for PCL/gelatin/HNT microfiber system loaded with metronidazole [44].

Conclusions
In this work, antibacterial, biodegradable, environmentally friendly, and nontoxic nanofibers with good biocompatibility and cost-effective production were prepared via the electrospinning of PCL and HNTs loaded with ERY. The PCL/HNT/ERY nanofibers exhibited outstanding antibacterial properties and resulted in the inhibition of both Gramnegative (Escherichia coli) and Gram-positive (Staphylococcus aureus) bacteria. We observed the gradual release of ERY out of the nanofibers, that was detected as repetitive antimicrobial activity after the material replacement on fresh prepared inoculated growth media. In addition to the antibacterial properties, the significant enhancement of mechanical properties was achieved via the incorporation of both unmodified and ERY-loaded HNTs, thus improving suitability of the system for medical applications. To summarize, PCL/HNT/ERY nanofibers with strong antibacterial effects have great potential as a progressive new biomedical material in both tissue engineering and a number of other biomedical applications.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.