Sciatic nerve repair using poly(ε‐caprolactone) tubular prosthesis associated with nanoparticles of carbon and graphene

Abstract Introduction Injuries to peripheral nerves generate disconnection between spinal neurons and the target organ. Due to retraction of the nerve stumps, end‐to‐end neurorrhaphy is usually unfeasible. In such cases, autologous grafts are widely used, nonetheless with some disadvantages, such as mismatching of donor nerve dimensions and formation of painful neuromas at the donor area. Tubulization, using bioresorbable polymers, can potentially replace nerve grafting, although improvements are still necessary. Among promising bioresorbable synthetic polymers, poly(l‐lactic acid) (PLLA) and poly(ε‐caprolactone) (PCL) are the most studied. Carbon nanotubes and graphene sheets have been proposed, however, as adjuvants to improve mechanical and regenerative properties of tubular prostheses. Thus, the present work evaluated nerve tubulization repair following association of PCL with nanoparticles of carbon (NPC) and graphene (NPG). Methods For that, adult Lewis rats were subjected to unilateral sciatic nerve tubulization and allowed to survive for up to 8 and 12 weeks postsurgery. Results Nanocomposites mechanical/chemical evaluation showed that nanoparticles do not alter PCL crystallinity, yet providing reinforcement of polymer matrix. Thus, there was a decrease in the enthalpy of melting when the mixture of PCL + NPC + NPG was used. Nanocomposites displayed positive changes in molecular mobility in the amorphous phase of the polymer. Also, the loss modulus (E”) and the glass transition exhibited highest values for PCL + NPC + NPG. Scanning electron microscopy analysis revealed that PCL + NPC + NPG prostheses showed improved cell adhesion as compared to PCL alone. Surgical procedures with PCL + NPC + NPG were facilitated due to improved flexibility of the prosthesis, resulting in better stump positioning accuracy. In turn, a twofold increased number of myelinated axons was found in such repaired nerves. Consistent with that, target muscle atrophy protection has been observed. Conclusion Overall, the present data show that nanocomposite PCL tubes facilitate nerve repair and result in a better regenerative outcome, what may, in turn, represent a new alternative to pure PCL or PLLA prostheses.

with some disadvantages, such as mismatching of donor nerve dimensions and formation of painful neuromas at the donor area. Tubulization, using bioresorbable polymers, can potentially replace nerve grafting, although improvements are still necessary. Among promising bioresorbable synthetic polymers, poly(l-lactic acid) (PLLA) and poly(εcaprolactone) (PCL) are the most studied. Carbon nanotubes and graphene sheets have been proposed, however, as adjuvants to improve mechanical and regenerative properties of tubular prostheses. Thus, the present work evaluated nerve tubulization repair following association of PCL with nanoparticles of carbon (NPC) and graphene (NPG).

Methods:
For that, adult Lewis rats were subjected to unilateral sciatic nerve tubulization and allowed to survive for up to 8 and 12 weeks postsurgery.
Results: Nanocomposites mechanical/chemical evaluation showed that nanoparticles do not alter PCL crystallinity, yet providing reinforcement of polymer matrix. Thus, there was a decrease in the enthalpy of melting when the mixture of PCL + NPC + NPG was used. Nanocomposites displayed positive changes in molecular mobility in the amorphous phase of the polymer. Also, the loss modulus (E") and the glass transition exhibited highest values for PCL + NPC + NPG. Scanning electron microscopy analysis revealed that PCL + NPC + NPG prostheses showed improved cell adhesion as compared to PCL alone. Surgical procedures with PCL + NPC + NPG were facilitated due to improved flexibility of the prosthesis, resulting in better stump positioning accuracy. In turn, a twofold increased number of myelinated axons was found in such repaired nerves. Consistent with that, target muscle atrophy protection has been observed.

Conclusion:
Overall, the present data show that nanocomposite PCL tubes facilitate nerve repair and result in a better regenerative outcome, what may, in turn, represent a new alternative to pure PCL or PLLA prostheses.

K E Y W O R D S
carbon nanotubes, graphene, PCL, peripheral nerves, tubulization

| INTRODUCTION
The peripheral nervous system has the remarkable capacity of regenerating following traumatic injury (Cunha, Panseri, & Antonini, 2011;Lewis, Courchet, & Polleux, 2013;Lopes et al., 2006). When there are significant gaps created by retraction or by extensive nerve damage resulting in loss of a nerve segment, grafts are required (Moroder et al., 2011). However, nerve autografts have some disadvantages, such as availability of donor tissue, the formation of painful neuromas in the donor area, and loss of donor nerve function (Bockelmann et al., 2011;Evans, 2000;Moroder et al., 2011;Nakamura et al., 2004;Pierucci, de Duek, & de Oliveira, 2008).
In this sense, many aspects of the tube used for nerve regeneration must be considered, for example, if it is bioresorbable, meaning that it undergoes degradation and the products and subproducts of the process are removed by metabolic pathways, causing no side effects.
Poly(ε-caprolactone) is a biomaterial that provides many of the desired properties for a tubing used for nerve regeneration. It is easy to handle and gives transparency, allowing better visualization of nerve stumps during surgery. It also provides flexibility, facilitating suturing of nerve stumps to the tubular prosthesis wall (Pierucci et al., 2008).
Nanometric supports have also been implemented to promote growth and development of neurons and non-neural cells (Mattson, Haddon, & Rao, 2000). The large class of nanomaterials currently available contains many families of nanostructures (Kotov et al., 2009).
Carbon nanostructures are the most evident products of nanotechnology and, among them, there are carbon nanotubes, carbon nanofibers, graphene, and a variety of carbon forms (Armentano, Dottori, Fortunati, Mattioli, & Kenny, 2010).
A carbon nanotube is characterized by rolling one or more graphene sheets concentrically with a diameter of nanometric dimensions and hollow inner cavity (de Oliveira, Ceragioli, & Assaf, 2014).
Carbon nanotubes are considered key materials and have many possibilities for technological applications due to properties such as high chemical resistance, oxidation and temperature resistance, low density, electrical conductivity, mechanical strength, and flexibility (de Oliveira et al., 2014). All these features make them excellent devices for neural implants, helping structurally and functionally for regeneration of damaged axons (Kotov et al., 2009;Sun, Sun, Li, & Peng, 2013).
Carbon nanotubes can also be excellent candidates for improving the properties of polymer composites due to features such as longitudinal elasticity, thermal and electrical conductivity, and low density (Terrones, 2013). When added to a polymer, they offer improvements in mechanical strength (Armentano et al., 2010). Liao et al. (2012) found that fibrous membranes of PLLA/PCL have benefits when added carbon nanotubes, such as increased mechanical properties, improved biodegradation, cell proliferation, and reorientation.
Graphene is also considered a promising material. Polymer nanocomposites filled with graphene have shown significant improvements in their properties such as tensile strength, electrical conductivity, and thermal stability (Zhu et al., 2010). In biomedical applications, graphene offers advantages to the nervous system, since the neural cells are electroactive and the electronic properties of graphene can be adapted for cargo transport required for cell electrical interface (Kotov et al., 2009).
Furthermore, chemical stability of graphene facilitates integration with neural tissues (Kotov et al., 2009). Li et al. (2011) observed that graphene films exhibit excellent biocompatibility in primary cultured rat hippocampal neurons and are also capable of promoting growth and sprouting of neurons, especially during the early development phase.
Therefore, the purpose of this study is to assess peripheral nerve regeneration, through transection of the sciatic nerve in rats, using tubular prostheses made of PCL alone or associated with nanoparticles of carbon (NPC) and graphene (NPG). Chemical and mechanical properties of the nanocomposites are also evaluated, demonstrating advantages of such associations.

| Construction of the prosthesis for tubulization
All biomaterial/composite membranes used to obtain tubular prostheses were prepared by a solvent evaporation technique as described previously (Pierucci et al., 2008). First, PCL was dissolved in chloroform (Merck, Germany) and kept under stirring for 2 hr for complete dissolution of the polymer. At the same time, solutions of carbon nanostructures were prepared in concentrations of 0.5% of NPC (multiwall), 0.5% of graphene oxide-NPG and 0.5% of the mixture of nanostructures (0.25% of NPC plus 0.25% of NPG) in chloroform.
Such solutions were placed in the ultrasonic bath for dispersion. NPC were prepared by HFCVD (Hot Filament Chemical Vapor Deposition) using acetone as carbon atoms source on a polished copper foil substrate as described previously (Grecco et al., 2011). The processing for NPG was the same described before to maintain identical properties (Mendonca et al., 2015). Briefly, NPG were prepared after catalytic conversion using a copper substratum to which 1 ml of polyaniline diluted in dimethylformamide (Synth, São Paulo, SP, Brazil) was added.
After drying for 2 hr at room temperature, 0.2 ml of nickel nitrate dissolved in pure acetone (Synth) was added to the preparation which was subsequently placed within a chemical vapor deposition reactor assisted by a hot filament. The hydrocarbons used as a carbon source were camphor and acetone.
The dispersed solutions of carbon nanostructures were mixed with the PCL solution and homogenized. These composites and PCL-alone solution were placed on glass molds into a glass vat for evaporation of the solvent and obtainment of the membranes.

| Fracture surfaces of the membranes
To evaluate the morphology, the fracture surfaces of pure PCL and nanocomposites were analyzed in a scanning electron microscope (Zeiss EVO MA15) at 10 kV. Fractures of the membranes were obtained after freezing in liquid nitrogen. The samples were metalized in a sputter coater (BAL-TEC SCD050) with gold.

| Differential scanning calorimetry
The differential scanning calorimetry (DSC) was performed on Netzsch equipment (DSC 200 F3 Maia, Germany). The samples were heated at 25-200°C, maintained for 5 min, after cooled down to −100°C, held for 5 min, and then heated at 100°C. Masses were used between 8 and 10 mg, heating rate of 10°C/min, a nitrogen atmosphere at 50 ml/min.

| Dynamic mechanical analysis
The dynamic mechanical analysis (DMA) was performed on Netzsch equipment (DMA 242). Samples were assessed at −100 to 60°C at a rate of 5°C/min, on the shear stress mode.

| Experimental groups
Forty adult female Lewis rats were obtained from the Multidisciplinary Center for Biological Investigation (CEMIB/UNICAMP) and housed under a 12-hr light/dark cycle with free access to food and water.
They were divided into four groups according to the composition of the implanted prosthesis: PCL (n = 10), PCL containing carbon nanotubes (NPC) (n = 10), PCL containing graphene oxide (NPG) (n = 10), and PCL containing carbon nanotubes and graphene oxide (a mixture of NPC + NPG) (n = 10). In each group, animals were sacrificed after a postsurgery survival period of 8 and 12 weeks (n = 5 per survival time).
All procedures were done in accordance with the ethical principles regulated by the National Council of Animal Experimentation (CONCEA) and with the approval of the Ethics Committee on Animal Experimentation of University of Campinas (CEUA/UNICAMP, protocol no. 2617-1).

| Specimen preparation
After 8 and 12 weeks postsurgery, the rats were perfused transcardially with phosphate-buffered saline (0.9% NaCl, PB 0.1 mol/L, pH 7.4) followed by Karnovsky solution (2% glutaraldehyde, 1% paraformaldehyde, PB 0.1 mol/L, pH 7.4). The contralateral and ipsilateral nerves were carefully dissected out under a surgical microscope. The ipsilateral nerve was dissected such that the proximal and distal stumps were separated. In sequence, the specimens were postfixed with 1% osmium tetroxide for 3 hr, washed with water, dehydrated with alcohol and acetone, and processed for resin embedding (Epon 812, EMS). Transverse semi-thin sections (0.5 μm thick) were obtained and stained with Sudan Black for morphometric analyses, counting of the regenerated fibers and measurement of the total nerve area.
Soleus and tibialis cranialis muscles were dissected out, on both sides, 12 weeks postsurgery (n = 3/group). The muscles were weighed for comparative analysis of their mass through the ipsilateral and contralateral ratio.

| Sciatic nerve area and counting of the regenerated fibers
Semi-thin transverse nerve sections were evaluated under light microscope (Leica DM5500B, Germany) camera and photographed with a high-resolution camera (Leica DFC345 FX). Regenerated nerves total transverse area was measured with Image J software (version 1.33u, National Institutes of Health-NIH, USA; RRID:SCR_003070) and to perform the counting of the regenerated fibers, 30% of the total nerve area was photographed at 1,000× magnification (Mayhew & Sharma, 1984a, 1984b. The total number of nerve fibers was estimated from a simple rule of three, considering the total nerve area.

| Scanning electron microscopy of the tubular prosthesis
The tubular prostheses were observed in a scanning electron microscope to evaluate their inner surface. For this, the samples were fixed with Karnovsky solution (2% glutaraldehyde, 1% paraformaldehyde, 0.1 mol/L PB, pH 7.4) and postfixed with 1% osmium tetroxide.
Following postfixation, the specimens were washed in distilled water and dehydrated in increasing alcohol series. Then, samples were dried at a critical point (Balzers CTD030), metalized in a sputter coater (BAL-TEC SCD050) with gold, and observed under a scanning electron microscope (Zeiss EVO MA15). Samples not implanted were used as controls.

| Statistical analysis
The data are presented as mean ± SE, and the differences between groups were considered significant when the p-value was <.05 (*), <.01 (**), and <.001 (***). Statistical analysis was performed with GraphPad Prism 5.0 software (RRID:SCR_002798). In this sense, data were subjected to ANOVA followed by Bonferroni post hoc test for parametric data or Mann-Whitney U test for nonparametric data.

| Fracture surfaces of the membranes
The  Figure 3 shows that there were no changes in the melting temperature (T m ) of the studied membranes. However, the enthalpy of melting (ΔH m ) increased after the addition of NPC and NPG to the first and second heating, whereas with the addition of the mixture of PCL + NPC + NPG there was a decrease in the enthalpy of melting.

| Differential scanning calorimetry
As seen in Table 1, T c , ΔH c , and χ c increase in pure PCL as compared to nanocomposites.

| Tubulization
After the tubulization of the sciatic nerve, the animals were observed during the postoperative periods, and there was no occurrence of self-mutilation, ulcers, or joint contractures at the site of injury.
All prostheses used in this study had transparency, contributing to the correct positioning of the nerve stumps when stitched at their ends. After 8 and 12 weeks of implantation, in the four studied groups, the tubes were found in the lesion site, connecting the ends of the nerves in the same way as were implanted. Moreover, in all analyzed tubes, a nervous cable was found connecting the proximal and distal stumps.

F I G U R E 3 Differential scanning calorimetry (DSC) curves. (a) First heating, (b) cooling, and (c) second heating
The materials used to manufacture the prosthesis shown to be biocompatible since no infection or inflammation in the lesion site could be depicted. Moreover, a thin layer of connective tissue formed on the inner and outer surfaces of the prostheses.

| Scanning electron microscopy of the tubular prosthesis
The inner surface of the tubular prosthesis was analyzed by scanning electron microscopy before implantation and after 8 and 12 weeks of surgery.
Before used for the implant, all membranes presented globular structures and pores, which are formed by solvent evaporation, ranging in size and number in different studied materials here ( Figure 5).
At the end of the studied periods, the globular structures and pores were rarely observed because there was deposition of extracellular matrix that overlaid the surface of the membranes. Moreover, the degradation of the material may have originated a flatter surface.

F I G U R E 4 Dynamic mechanical analysis (DMA) curves. (a) E' (storage modulus) and (b) E" (loss modulus)
adherent red blood cells and rarely appeared another cell type. Just scanning electron microscopy (SEM) did not allow us to identify the cell types present there. Red blood cells could be recognized because they have a very distinctive shape. At 12 weeks, the deposited matrix was thicker and covered a larger area. Also, the amount of adhered cells was lower as compared to 8 weeks post implantation. In both tubes, 8 and 12 weeks postoperative, structures were seen suggesting the presence of blood vessels during tissue regeneration.
After 8 weeks of implantation, the NPC containing membranes also showed many adherent cells on its surface and, unlike the PCLalone group, with a smaller amount of red blood cells. Furthermore, there was a significant amount of matrix deposited on such nanocomposite membranes. After 12 weeks, the number of cells and deposited matrix were less prominent when compared to the previous period.
Also, it was possible to observe structures indicating the presence of blood vessels as well as signs of degradation of the polymer.
Membranes prepared with the mixture of PCL + NTC + NPG, after 8 weeks, displayed less prominent matrix deposition as compared to 12 weeks period. Adhered cells, sending cytoplasmic projections, were compatible with the morphology of fibroblasts. Of note, polymer degradation/reabsorption appeared significantly more advanced in this group, 12 weeks post-tubulization, presenting a large number of fissures and infiltrating cells, possibly macrophages.

| Counting of regenerated fibers
After 8 and 12 weeks, the number of fibers in the regenerated sciatic nerve of each animal was also estimated. 8,381 ± 2,158; mean ± SE; Figure 7n).

| Ratio of the masses
There was no statistical difference between groups (p > .05) when comparing the ratio of the masses of the soleus muscle at ipsilateral and contralateral sides after 12 weeks of nerve repair. The same was PCL + NPC + NPG. After 12 weeks, matrix deposition was greater, but with few cells. Magnification: 800×. P, area not covered by matrix; M, area covered by matrix; arrow, cells; asterisk, marks indicative of blood vessels. Scale bar: 10 μm, 1,500×. PCL, poly(ε-caprolactone); NPC, nanoparticles of carbon; NPG, nanoparticles of graphene stimulate axonal growth by facilitating cell migration (Mattson et al., 2000). Herein, we show that the combination of PCL + NPC + NPG produces homogeneous membranes that can be rolled into tubular prostheses and used for tubulization of sciatic nerve post-transection (neurotomy). In a previous study, we have already demonstrated that PCL alone allows proper nerve regrowth and successful reinnervation of hind limb musculature (Pierucci et al., 2008). By using the solvent evaporation approach, thin and resistant tubes are produced, and the combination with NPC + NPG improved significantly chemical and mechanical properties of the composite. as seen by MEV. This is consistent with the concept that a nerve con-  (Sun et al., 2013). In turn, the combination of NPC + NPG produced the best results, taking advantage of positive characteristics of both nanoparticles.
Cell proliferation and migration to the inner tube space is of fundamental importance for a successful regenerative process (Dubey, Letourneau, & Tranquillo, 1999;Geuna et al., 2009). It is known that such space is quickly filled with plasma exudate composed mostly of a fibrin network and platelets. It has been recently shown that blood vessel formation and polarized growth precedes Schwann cell migration, which in turn produce the so-called Bands of Büngner (Cattin et al., 2015). Such cell cords guide the growing axons toward the distal stump of the nerve.
Importantly, implanted tubular prostheses are enclosed by connective tissue both at the outer and inner wall surfaces. In this sense, the use of hydrophilic biomaterials may result in thick encapsulation and, in turn, compression of the regrowing nerve. This can be avoided by using PCL, which is hydrophobic (Subramanian, Krishnan, & Sethuraman, 2009). Nevertheless, the balance provided by mixing PCL with NPC and NPG results in a biofunctional nanocomposite that allowed only a thin layer of tissue covering, what possibly facilitated the regenerative process. In this regard, nanostructures of carbon, due to their negative charge, may facilitate the initial steps of cell adhesion and proliferation (Bergethon, Trinkaus-Randall, & Franzblau, 1989;Lee, Cuddihy, & Kotov, 2008;Oliveira et al., 2014;Zhu et al., 2010). As a result, significant larger nerve bridges were found within PCL + NPC + NPG tubes, which presented greater number of myelinated axons.
One key point for the success of nerve tubulization by using bioresorbable polymers is the composite degradation rate (Moroder et al., 2011). As mentioned earlier, PCL alone presented larger globular structures as compared to the nanocomposites. In this sense, the increased surface of contact possibly facilitated hydrolysis and the initial steps of degradation, generating fissures and cavities already at 12 weeks postsurgery. Nonetheless, at that stage the nerve cable is well formed and stable, so that reabsorption of the composite should not interfere in the functional motor/sensory outcome. The process of hydrolysis, however, if too fast, may generate oxidative stress and a local decrease in pH (de Ruiter et al., 2009). Based on the fact that muscle mass was preserved in all tubulized groups, no harmful effect of the prosthesis degradation should have affected the newly formed nerve (Dow et al., 2004;Yu et al., 2011).
In conclusion, the present work indicates that combining PCL with nanoparticles of carbon and graphene is a promising approach, since such association improves chemical/mechanical aspects of the composite, even accelerating its degradation rate. Also, biocompatibility is enhanced, as NPC and NPG are negatively charged, facilitating cell migration into the tube. Overall, nerve regeneration was more prominent when PCL + NPC + NPG was used, generating thicker nerve cables, containing increased number of myelinated axons.