A multi‐layered nerve guidance conduit design adapted to facilitate surgical implantation

Abstract Background and Aims The gold standard procedure after a severe nerve injury is the nerve autograft, yet this technique has drawbacks. In recent years, progress has been made in the development of artificial nerve guides to replace the autograft, but no device has been able to demonstrate superiority. The present study introduces an adaptable foundation design for peripheral nerve regeneration. Methods Silk fibroin was electrospun, creating a tri‐layered material with aligned fiber surfaces and a randomly deposited fiber interior. This material was rolled into a micro‐channeled conduit, which was then enveloped by a jacket layer of the same tri‐layered material. Results The proposed implant design succeeds in incorporating various desirable aspects of synthetic nerve guides, while facilitating the surgical implantation process for medical application. The aligned fiber surfaces of the conduit support axon guidance, while the tri‐layered architecture improves its structural integrity compared with a fully aligned fiber material. Moreover, the jacket layer creates a small niche on each end which facilitates surgical implantation. An in vivo study in rats showed that nerve regeneration using this device was comparable to results after direct suture. Conclusion This proof‐of‐principle study, therefore, advances the development of tissue engineered nerve grafts by creating an optimized guidance conduit design capable of successful nerve regeneration.

After a severe injury to a peripheral nerve, the axon segments distal to the trauma site begin to degrade, which results in a local loss of function as organs are denervated, during a process called Wallerian degeneration. 3,4 Once the distal axon segments degrade, the proximal axon stubs begin to regrow towards their respective innervation targets, assisted by biochemical cues and structural guidance from proliferating Schwann cells. 3,5 In many minor nerve injury cases, nerves will regenerate naturally, resulting in functional recovery without the need for surgical intervention. However, for more serious injuries, where the gap between uninjured nerve segments cannot achieve a tension-free coaptation, a nerve graft must be implanted in order to bridge the gap between the proximal and distal segments of the nerve. 6 In North America, between roughly 50 000 and 200 000 surgeries for nerve repair are performed each year. 7 The current gold standard for nerve repair when direct end-to-end suture is impossible, is the nerve autograft, which consists of grafting the damaged nerve with a nerve sample taken directly from the patient; samples are most often taken from the sural nerve located in the leg and the medial antebrachial cutaneous nerve located in the arm. 8 Other sensory nerve harvesting locations may be considered by surgeons, such as the superficial cervical plexus. 9 Not only is this method's effectiveness in nerve regeneration limited, resulting in full functional recovery in only half of cases, but it also has several other drawbacks, such as the need for multiple operations, limited donor tissue availability, loss of function at the donor site, and the risk of developing a painful neuroma at the donor site. 5,10 Before following through with a nerve autograft procedure, it is important that the surgeon informs the patient of the consequential postoperative sensory loss in either the lateral portion of the foot or the medial portion of the mid-forearm (a common side effect of an autograft from the sural nerve and medial antebrachial cutaneous nerve, respectively), as the loss of sensory function in one area may be more or less significant for each patient.
In order to overcome the drawbacks presented by the autograft, researchers have turned to tissue engineering in order to develop a more effective alternative, using biomaterials. 11 In particular, the natural polymer silk fibroin is a promising material for the fabrication of a nerve guidance conduit, since it is biocompatible, biodegradable, easily functionalized, easily chemically modified, and has robust mechanical properties compared with other natural materials. 12 In addition to the choice of biomaterials, an effective nerve guidance conduit must provide an environment that encourages healthy, guided axon regrowth. 5 The conduit must, therefore, succeed in preventing the regrowing axons from straying far from their original paths towards their innervation targets. If regenerated neurons were misdirected during regrowth, not only will functional recovery be significantly reduced in intensity, but the patient may also require sensory and motor re-education therapy in order for the brain to recognize an entirely new input, as neurons that once innervated one area of the body may have reinnervated a different area. 8 In cases of minor injuries to the nerve, axons are directed by the naturally realigning Schwann cells that release biological cues, thus attracting growth cones and successfully leading the axons along their original paths towards their innervation targets. 13 The autograft has an advantage regarding guidance capabilities due to the numerous micro-channels in the predeveloped fascicles which increase the surface area to volume ratio and lower the possibility of axons straying from a predetermined route. 14,15 Most artificial nerve guidance conduits on the market, however, are hollow tubes that do not provide as much mechanical support. 14 The epineurium is the outermost layer of the nerve and is made up of dense connective tissue which houses the nerve fascicle; these have their own layer of protective connective tissue called the perineurium.
In most cases, autografts or nerve guidance conduits are sutured at the epineurium of the nerve at the proximal and distal nerve stumps in order to connect a severed nerve. 15 In some cases, several nerve grafts are sutured directly at the perineurium of each fascicle with the aim of better fascicle matching. Micro-suture at the perineurium, however, results in increased trauma and scarring to the nerve, and there is no consistent evidence of superior results to epineurial sutures. 6,15 However, after a nerve is severed, surgeons observe that the epineurium will naturally retract a small percentage further than the nerve fascicles due to a release in tension. This consequence poses a challenge for surgeons to correctly position implants with blunt edges, such as an autologous nerve graft. This is because as the epineurium is stretched to connect with the outer layer of a blunt-edged implant, the fascicles may become deformed, weakening the prospect of successful neuron guidance. Hollow nerve guidance conduits, such as Neuroflex or NeuraGen, overcome this obstacle, since the protruding nerve fascicles are simply inserted into the hollow cavity of the implant. 16 However, hollow conduits provide minimal directionality due to lower physical support.
Finally, the mechanical strength of the device is an important parameter that must be considered. 17 The implant must be resistant to all forces acting upon the implant during and after implantation. For example, a weak resistance to tensile forces after surgery can lead to poor regeneration, and surgeons request that regenerating neurons travel a longer distance through a longer implant, as opposed to a shorter implant that is put under excessive stress. 8,15 Therefore, the implant must uphold a certain stability while under tensile stress in order to achieve a conduit with a relatively small length and an environment that will promote healthy neuron growth. The material must also withstand the trauma of suture during the surgery and, therefore, must possess strong tear strength in order to facilitate successful implantation.
Silk fibroin extracted from the Bombys mori silk worm cocoons was chosen for this study because of its numerous advantageous properties as a natural material. Once the fibroin protein is purified from the raw silk cocoons, it is a biocompatible material that generates a weaker inflammatory response than that of both collagen and PLA, which are commonly investigated biomaterials for nerve guidance conduit fabrication. 5,12 Silk fibroin is an interesting biomaterial for this study also because it is easily chemically modifiable as well as functionalizable with diverse substances 18 ; material functionalization could ultimately be optimized to yield a superior biomaterial complex.
In addition, the degradation properties of silk fibroin can be controlled during material fabrication. Hu et al demonstrated that increasing the amount of β-sheets in the protein secondary structure ultimately slows biodegradation. 19 Finally, silk fibroin has already been FDA approved as a biological suture material.
The design of the device presented in this study takes several factors into consideration. First, the material and material structure were chosen for biocompatibility, versatility, and mechanical integrity. Silk fibroin was electrospun to create a complex, tri-layered nanofiber material optimizing parameters to allow both surface alignment and good mechanical strength. Second, micro-channels were included in the fabrication of the nerve guidance conduit in order to incorporate a significant advantage valued in the nerve autograft. Finally, a jacket layer was added to the multi-channeled conduit in order to incorporate the principal advantage to hollow nerve guides, which is to facilitate the surgical procedure by allowing a more straightforward epineurial micro-suture technique. Therefore, the goal of this study was to develop an adaptable implant foundation design capable of providing enhanced guidance to regenerating neurons that also caters to the needs of the surgeon during implantation.

| Preparation of silk fibroin solution
A 10 wt% silk fibroin solution was obtained using a previously established protocol. 33 Briefly, silk cocoons from the Bombyx mori silkworm were cut into small pieces and boiled for 30 minutes in a 0.02 M Na 2 CO 3 aqueous solution. The silk fibroin fibers were rinsed three times in DI water and then allowed to dry at room temperature. The dry fibroin fibers were dissolved in a 9.  RPM. To obtain a randomly deposited fiber material, the spinning solution was dispensed continuously for 90 minutes with a collector rotation speed of 400 RPM. Aluminum foil was used to cover the collector surface before each process to allow for easy sample recovery.

| Implant fabrication-jacketed, multi-channel design
The implant fabrication process is shown in Figure S1. The tri-layered electrospun material was carefully peeled from the collector, and a 5 mm (parallel to aligned fibers) by 3 cm (perpendicular to aligned fibers) rectangle of the material was cut using a scalpel blade. The material was then rolled while adding a Teflon-coated stick (0.2-mm diameter) after every full rotation. Once completely rolled, the tube was immediately immersed in methanol for 5 minutes to induce βsheet formation. The tube was allowed to air-dry for 1 hour, and the Teflon-coated sticks were then removed, resulting in a 5-mm-long multi-channeled tube.
From the tri-layered electrospun material, a 7 mm (parallel to aligned fibers) by 3 cm (perpendicular to aligned fibers) rectangle was cut. The multi-channel tube was placed at the bottom-center of this rectangle (aligned fibers in the same direction). The larger material was then rolled around the tube for 3 rotations to create a "jacket." The edge was tightly pressed to the tube, and the entire device was water vapor annealed at room temperature for 4 hours to induce β-sheet formation. The implants were immersed in Milli-Q water overnight at 37°C and then rinsed three times in order to eliminate traces of PEO. The implants were then sterilized in 70% ethanol overnight, rinsed three times with sterile water, and immersed in sterile PBS for storage.

| Mechanical strength testing
For tensile strength tests, purely aligned, randomly deposited, and trilayered electrospun material samples were rolled four rotations to produce a hollow tube. The tubes were water vapor annealed for 4 hours at room temperature and then immersed in Milli-Q water overnight at 37°C to extract the PEO from the fibers. The tubes were subsequently rinsed and then immersed in PBS prior to testing.
Hydrated material samples were consecutively secured lengthwise between the upper and lower holding grips with a gauge length of 3 mm. Each trial was carried out at a cross-head speed of 0.06 mm·s −1 while recording load measurements every 100 ms until rupture.
Assays for each material were done in triplicate. All values are represented by mean ± standard deviation.
In order to test the tear strength of the materials, aligned and trilayered material samples wear punctured with a 9-0 round bodied suture needle with a polyamide 6/6 thread 1 mm from the materials' edge. The suture thread was trimmed on both sides of the puncture, but not knotted. The threaded material was secured in the lower holding grips at the opposite edge of the puncture, while the two free suture threads were secured in the upper holding grips. The sutured material was adjusted automatically in order to assure equal tension between both sides of the suture thread. The tensile force of the system was then measured with a cross-head speed of 0.06 mm·s −1 up to a maximum displacement of 4 mm. Assays were done in triplicate. All values are represented by mean ± standard deviation.

| Fiber diameter and angle analysis
From SEM images at randomly chosen areas of the aligned material surface, 100 fiber diameters were measured, and 50 fiber angles were measured using ImageJ image analysis software (version 1.50i). Diameters of each fiber were measured at the center of the SEM image unless otherwise hidden from view at the image's center. Fiber angles were measured by drawing straight lines along the fiber and the y axis of the image. All angles are expressed relative to the primary alignment of the aligned fibers (0°), which was calculated by subtracting the average fiber angle from the measured fiber angle relative to the y axis of the image. All calculated fiber angles were from a single SEM image. All values are represented by mean ± standard deviation.

| In vitro study
Electrospun silk fibroin fibers and glass coverslips were immersed in 70% ethanol for 24 hours and allowed to dry prior to cell seeding.

| Surgery
Twelve 6-week-old, male Sprague Dawley rats were each anesthetized with Vetflurane (4.5% during induction and 3.5% during surgery, Virbac) throughout the entirety of the operation. The rats' right hind limbs were each shaved and sterilized. For each rat, an incision parallel to the femur was made, and the sciatic nerve was exposed, isolated, and fixed with two micro clips 10 mm apart. The nerve was severed with surgical scissors at two points 5 mm apart between the micro clips, and the extracted portion of the nerve was discarded.
For the implantation of nerve guidance conduits made from aligned fiber materials without jacket layers (four animals), the edge of the implant was sutured (Ethicon Ethilon Polyamide 6/6 suture, 9-0 round bodied) at four points to the distal segment's epineurium (at 0°, 90°, 180°, and 270°). The proximal nerve segment was then lined up with implant's proximal end. The edge of the implant was subsequently sutured at four points to the proximal segment's epineurium (at 0°, 90°, 180°, and 270°).
For the implantation of nerve guidance conduits made from the tri-layered fiber material with a jacket layer (four animals), the implant outer layer was sutured (Ethicon Ethilon Polyamide 6/6 suture, 9-0 round bodied) twice to the distal segment's epineurium (at 0°and 180°) enveloping the epineurium and nerves fascicles. The exposed epineurium and fascicles of the proximal nerve segment's severed edge were inserted into the small cavity at the proximal end of the implant, and the implant outer layer was sutured twice to the epineurium (at 0°and 180°). The operated area was then cleaned, and the wound was closed and sutured.
After suture, the operated areas were cleaned, and the wound was closed and sutured. Two experimental groups plus a control group were studied in an 8-month in vivo study on nerve regeneration of the sciatic nerve in the Sprague Dawley rat. Experimental group 1 included four animals with a jacketed silk fibroin-based nerve guidance conduit implanted between two severed sciatic nerve segments, as previously described in the Surgery section. Two animals were analyzed at month 4 of the study, and two animals were analyzed at month 8 of the study.

| In vivo study
Experimental group 2 included four animals whose sciatic nerve in the right hind limb was severed as previously described in the Surgery section and subsequently secured end-to-end with four micro-sutures at 0°, 90°, 180°, and 270°. Two animals were analyzed at month 4 of the study, and two animals were analyzed at month 8 of the study.
The operated area was then cleaned, and the wound was closed and sutured.
The control group animal was not operated on.
Half of the animals from each experimental group were euthanized at 4 months after surgery and half at 8 month after surgery. Segments of the sciatic nerve cut within 10 mm distal and proximal to the site of injury were harvested from each animal prior to euthanization and immersed in 4% formaldehyde for at least 3 days.

| Tissue sectioning
Nerve segments were immersed in subsequent baths of 15% sucrose solution and 30% sucrose solution then flash frozen in OCT compound (Fisher Healthcare, 23-730-571). Nerve sections with a thickness of 8 μm were collected using a cryomicrotome (Leica CM 3050 S).
Tissue sections were immunostained either with a combination of DAPI, anti-β-tubulin III (Sigma Aldrich, T2200), and Alexa Fluor 488 phalloidin for visualization of nucleic acids, β-tubulin III, and actin filaments respectively, or a combination of DAPI, anti-β-tubulin III, and anti-myelin protein zero (Abcam, Ab134439) for visualization of nucleic acids, β-tubulin III, and protein zero, respectively. Immunostaining procedures were carried out according to the manufacturers' instructions.
Samples were visualized with either an epifluorescence microscope (Leica DMI 6000B, 20x/0.40) or a confocal microscope (Zeiss LSM 710, 63x/1.40 Oil DIC) using z-stacking method. Fifty nerve fiber diameters and 50 myelin sheath thicknesses in each sample were measured using ImageJ image analysis software (version 1.50i). All values are represented by mean ± standard deviation.

| Nerve guidance conduit design
After a total of 3 × 30 minutes of electrospinning, a tri-layered fibrous silk fibroin material 3 cm wide and 80 μm (±10 μm) thick was obtained.
The tri-layered material consisted of an internal layer of randomly deposited fibers, which was sandwiched between two layers of aligned fibers. Fiber diameters were found to be 417 ± 134 nm, with 94% of diameters within the range of 200 nm and 600 nm. In the aligned fiber layers, 86% of fiber angles were found to be within ±5°o f the primary alignment. The tri-layered material exhibited three distinct layers visible through SEM analyses (Figure 1). The aligned-fiber surfaces of the tri-layered material were also shown to completely cover the random fiber center layer ( Figure 1B), assuring that the guidance factor of this material was not compromised.
A multi-channeled conduit capable of maintaining its structure after water vapor annealing (necessary to induce the formation of silk fibroin β-sheets) was obtained and is shown in Figure 2. The tri-layered material samples preserved continuity long after initial rupture ( Figure 5).

| Mechanical tests
Tear strength characterizations of both aligned and tri-layered materials are shown in Figure 6. The sutured tri-layered material was able to resist an average maximum force of 50.7 mN ± 2.0 mN, which was reached at a displacement of 2.2 mm ± 0.2 mm. The aligned

| In vitro study
After Wallerian degeneration, Schwann cells must be able to migrate through the nerve guidance conduit in order to create paths towards which the regeneration nerve fibers will be guided. Therefore, rat Schwann cells were seeded on the aligned silk fibroin fiber material surface in order to analyze cell behavior on this material in vitro.
Compared with control samples consisting of rat Schwann cells seeded on glass cover slips, Schwann cells appeared to thrive similarly on each sample surface yet exhibiting morphological differences (see

| Surgery
As a preliminary test to simulate the procedure carried out by a surgeon when performing a nerve repair intervention and to evaluate the surgical handling of the guide, surgery was carried out on the right sciatic nerve of 12 male Sprague Dawley rats. Implantation of the trilayered jacketed nerve guidance conduit is depicted in Figure 8 The aligned nerve guidance conduit without a jacket layer was sutured between severed segments of the sciatic nerve and is pictured in Figure 9A. Due to the blunt edge of this implant, the nerve fascicles had limited space during the epineurial suture. In consequence, the nerve segment edges were slightly deformed while performing the micro-suture.

| In vivo study
A direct end-to-end suture is the preferred method for repairing nerves. Only when the gap between nerve segments is too large and end-to-end suture will put excessive tensile stress on the nerve would a graft be a more suitable repair technique. 8 Sciatic nerve sections from an area distal to the site of injury were taken after 4 months or 8 months in vivo from nerves repaired with either a direct suture or a silk fibroin-based nerve guidance conduit. All nerve sections were taken within 10 mm distal to the most distal suture at the site of injury or from an uninjured sciatic nerve.
Samples taken from experimental groups after 4 months in vivo, plus a control sample taken from an uninjured sciatic nerve, were immunostained for visualization of regenerated nerve fibers FIGURE 9 A, Aligned, multi-channeled silk fibroin nerve guidance conduit sutured between two sciatic nerve segments. B, Tri-layered, multichanneled silk fibroin nerve guidance conduit with jacket layer sutured between two sciatic nerve segments FIGURE 8 Implantation procedure: A, the rat sciatic nerve was exposed and secured (dashed lines indicate where the nerve was severed with surgical scissors; the portion of nerve between dashed lines was extracted); B, the implant was subsequently sutured once at 0°and once at 180°a t the distal nerve segment; C, the proximal nerve segment was placed inside the implant cavity; D, the implant was sutured once at 0°and once at 180°at the proximal nerve segment. Blue arrows indicate the point at which the device was sutured to the nerve epineurium. The green arrow highlights the secured nerve before suture to the device ( Figure 10). In all samples, abundant neuron regeneration was seen in direct suture nerves as well as in nerves having received a nerve guidance conduit. In addition, samples from both experimental groups showed a homogenous distribution of nerve fibers throughout the nerve fascicles similar to the control sample.
After 8 months in vivo, both experimental groups showed improvement in nerve regeneration compared with the results obtained at 4 months, as seen in Figure 11.  both at the proximal and distal nerve segments. A multi-channeled conduit, however, has blunt edges and risks the deformation of nerve fascicles, as the outer layer of the conduit is sutured directly to the edge of the nerve segments' epineurium. After the nerve is sectioned, the epineurium retracts leaving protruding nerve fascicles, making fascicle deformation virtually impossible to avoid. Responding to this concern, the presented implant design features a "jacket" layer enveloping the guidance conduit. The jacket, which is 1 mm longer than the multichanneled conduit on each end, presents a hollow cavity for the place- The significant increase in ductility of the tri-layer material is represented by the necking observed from the tensile strength tests and the continuity of the tri-layer material that was kept long after reaching ultimate strength. In contrast, the aligned fiber material met complete or quasi-complete rupture immediately after ultimate strength was reached. The increased ductility, together with a comparable resistance to deformation, is a benefit for a material used to create a nerve guide in order to ensure that the implant will not fail if the implant experiences an exceptionally large load. Despite the slight decrease in tensile strength with the addition of the center random fiber layer, the tri-layer material has a greater capability than the aligned fiber material to keep continuity after experiencing a similar load that reaches or exceeds slightly the maximum tensile stress of either material.
An additional mechanical property that needs to be addressed for a nerve guide is the material's resistance to tearing. Nerve guides are often sutured to the epineurium during the surgical procedure, which requires a puncture to the material very close to the material edge. 15 If the material has a poor resistance to tearing, a small load from the suture thread after puncturing the material will cause the implant to detach from either nerve segment, and the implant will fail. The trilayer material was shown to possess a much higher resistance to tearing than the aligned material. With the addition of the center random fiber layer, the maximum force withstood from the sutured tri-layer material increased by 371.6%. In addition, the elongation of the sutured tri-layer material at its maximum force was 734.5% higher than that of the aligned material. There are two main benefits of this significant increase of the maximum force and elongation withstood by the tri-layer material. The first, is an increased resistance to tearing, which will more successfully prevent the implant from detaching from the nerve segments in the weeks following the surgery. The second benefit is the increased ease for surgeons to successfully suture the device while devoting less attention to the force being applied to the device during suture.
This novel implant's design application efficacy was compared with a blunt-edged multi-channel design using purely aligned silk fiber material. All surgeons participating in this work highly preferred the Contrarily, random positioning of Schwann cells throughout the guidance conduit could cause the nerve fibers to take different paths leading to incorrect tissues.
In order to evaluate the success of nerve regeneration using the presented device, an in vivo study using Sprague Dawley rats was per- This study emphasizes the importance of the design of a nerve guidance conduit including the choice of biomaterial used, the material fabrication technique and organization, and the architecture of the three-dimensional device itself. This foundational design for a nerve guidance conduit has shown to promote nerve regeneration success comparable to that of an end-to-end suture at the rat sciatic nerve long term. Histological results revealed that the main difference between this nerve guidance conduit and the end-to-end suture interventions was a slightly delayed restoration of nerve fiber diameter size and myelin sheath thickness, signifying a slightly delayed regeneration.
While possible explanations for these results have previously been expressed, encouraging a more accelerated regeneration time may be explored by expanding upon the nerve guidance conduit model presented in this study. For example, the functionalization of the material with growth factors may promote accelerated nerve fiber extension. Growth factors that have already been explored to promote enhanced nerve regeneration include nerve growth factor, ciliary neurotrophic factor, neurotrophin-3, and glial derived neurotrophic factor. 5 For example, Wang et al found more regenerated fibers and a larger gastrocnemius muscle weight ratio from poly (DL-lactide-coglycolide) nanofiber nerve guidance conduits functionalized with nerve growth factor compared with non-functionalized poly (DLlactide-co-glycolide) conduits. 30

| CONCLUSION
The jacketed, micro-channeled conduit based on a nanofibrous material organized in a tri-layered architecture presented in this work is a base design for a nerve guidance conduit that addresses the surgical application concerns that are regularly overlooked in the literature.
As a result of the micro-channels, the device exhibits a high surface area to volume ratio providing an aligned nanofiber surface in order to increase axon guidance during regeneration. Furthermore, the conduit contains a tri-layered jacket, which significantly increases the ease of surgical implantation. Histology analyses throughout a long-term in vivo study showed that nerve regeneration using this nerve guidance conduit was comparable to regeneration results from an endto-end suture, which is the preferred technique of nerve repair if the gap between severed nerve segments is minimal. In addition, the device may be adapted further to ameliorate nerve regrowth efficacy by modifications of the material or bio-functionalization. Silk fibroin was used for this model because of the advantageous properties for biological devices. Silk fibroin is a natural, biocompatible polymer with robust mechanical properties that is biodegradable, easily chemically modifiable, and easily functionalized. Therefore, silk fibroin supports the incorporation of many functionalizing materials or bio-signaling molecules in order to improve the performance of the device.