Abstract
This is a proof of principle report showing that fibres of Bioglass® 45S5 can form a biocompatible scaffold to guide regrowing peripheral axons in vivo. We demonstrate that cultured rat Schwann cells and fibroblasts grow on Bioglass® fibres in vitro using SEM and immunohistochemistry, and provide qualitative and quantitative evidence of axonal regeneration through a Silastic conduit filled with Bioglass® fibres in vivo (across a 0.5 cm interstump gap in the sciatic nerves of adult rats). Axonal regrowth at 4 weeks is indistinguishable from that which occurs across an autograft. Bioglass® fibres are not only biocompatible and bioresorbable, which are absolute requirements of successful devices, but are also amenable to bioengineering, and therefore have the potential for use in the most challenging clinical cases, where there are long inter-stump gaps to be bridged.
INTRODUCTION
Repairing a traumatically injured peripheral nerve can be a protracted and frequently incomplete process; the clinical outcome is often unsatisfactory and there is rarely a complete return of function (Birch et al., 1998; Lundborg, 2000). When severed nerve ends cannot be reapposed by tension-free sutures, these injuries are conventionally managed using an interfascicular or group fascicular autograft. The latter contain acutely denervated, “axon-responsive” Schwann cells, lying within a scaffold of longitudinally aligned basal laminae, and provide a microenvironment that is known to facilitate axonal regrowth (Hall, in press). However, although they remain the gold standard, autografts offer a less than ideal solution to the problem of bridging an inter-stump gap, particularly when the tissue defect is extensive (Shirley et al., 1996). Complications associated with their use include limited tissue volume and fascicular and modality mismatches between host and donor nerves, secondary donor site morbidity, neuroma formation and numbness within the distribution of the donor nerve.
The continuing search for an alternative to a nerve autograft has taken two main routes. In broad terms, either the nerve stumps are enclosed (entubulated) within some type of non-neural tube, fashioned from natural or synthetic material (McDonald and Zochodne, 2003) or a suitably prepared peripheral nerve allograft, used in combination with low-grade immunosuppression, is implanted into the inter-stump gap (Grand et al., 2002; Udina et al., 2004). Thus far, none of these options has outperformed an autograft when used in a short gap, and none has proved effective in bridging long gaps.
An extensive literature search suggests that the design specifications for an alternative to a nerve graft (whether autograft or allograft) must include the provision of a biocompatible and bioresorbable scaffold that supports outgrowing axons and their associated cells and within which the microenvironment of a peripheral nerve fibre can be replicated. Very similar generic requirements have driven the search for alternatives to autografts in replacing bone defects: research in this field has moved with speed in the last decade and has benefited greatly from the development and manipulation of bioactive glasses and ceramics. Bioglass® is a widely used melt-derived bioactive glass (Hench, 1998; Hench et al., 2004; Kaufmann et al., 2000; Loty et al., 2001; Maquet et al., 2004; Verrier et al., 2004; Xynos et al., 2000, 2001) and its surface is chemically active. Biomaterials have been developed that take advantage of the ability to tailor its surface to specific applications (Lu et al., 2003).
We report here, in a proof of principle study, that fibres of Bioglass® entubulated within silastic conduits provide a biocompatible and bioresorbable scaffold that supports axonal regeneration across 0.5 cm gaps in the sciatic nerves of adult rats: the axonal regrowth is qualitatively and quantitatively indistinguishable from that seen using an autograft.
MATERIALS AND METHODS
Bioglass 45S5 fibre preparation
Bioglass 45S5 (45% SiO2, 24% Na2O, 24.5% CaO and 6% P2O5 in weight percent) was prepared by melting in a BLF-1700 high-temperature furnace at 1250 and 1350 °C for 45 minutes each. The platinum crucible was removed from the furnace, and placed on a thermo-resistant asbestos plate. A Bioglass 45S5 monolith rod of 10 mm diameter was immersed in the liquid bioglass and removed at a rate of approximately 1 metre/second. The attached viscous bioglass 45S5 formed a continuous fibre, whose diameter could be altered by varying the removal speed. The resulting solid fibres were cut into 10 cm-long sections, visually inspected at × 20 magnification for imperfections such as inclusions and stress cracks and then sorted into diameter sizes by microscopical sorting on a micrometer grid. Fibres were washed twice using 70% methanol and allowed to air dry under a stream of sterile air before use. Aseptic procedures were used during subsequent handling of the fibres to ensure that they remained sterile.
Interactions of Bioglass® with cells in vitro
Preparation of Schwann cells
Male Wistar rats (225 g) were killed by CO2 asphyxiation. Their sciatic nerves were exposed in mid-thigh through a muscle-splitting incision, and all major branches of the common peroneal and tibial nerves were dissected out and washed twice in Hank’s Balanced Salt Solution (GIBCO) supplemented with gentamycin (50 μg/ml; Sigma Chemical Co., St. Louis, MO). The epineurium and perineurium were stripped off with fine forceps under a dissecting microscope. Desheathed nerves were blotted with sterile filter paper, chopped into 1 mm segments using a McIlwain tissue chopper and placed in 35 mm dishes containing Dulbecco’s Modified Eagle’s Medium (DMEM) (GIBCO, Grand Island, NY), supplemented with 10% fetal calf serum (FCS), (DMEM/FCS), and 50 μg/ml gentamycin. Explants were left floating in this medium for 4 days at 37 °C. Collagenase/dispase (0.1%, Boehringer Mannheim Biochemicals, Indianapolis, IN) was added to the medium 18 hours before dissociation, and explants were dissociated on day 4 by gentle trituration through a flame-narrowed pipette (0.5–1 mm bore). Cells were stained with trypan blue (to differentiate them from myelin debris) and the number of viable cells was counted using a haemocytometer. To determine the proportions of the constituent cell types, aliquots containing 104 cells were plated onto poly-
Determination of glass fibre biocompatibility
Five fibres (∼10 mm in length; ∼150–200 μm diameter) were placed into each well of a 24-well cell culture plate. One millilitre of DMEM/FCS containing 2 × 104 cells was added to each well. After plating for 24 hours, the culture medium was removed from each well and the cells fixed with 250 μl of either 4% PFA for 1 hour. Several glass fibres fixed with PFA were stained with toluidine blue, mounted in aqueous mounting medium (Serotec) and inspected. The remainder were immunostained with anti-S100 (DAKO; 1:100; 1 hour at RT) which was visualized with anti-rabbit FITC (Sigma; 1:100; 1 hour at RT). Fibres were mounted in Citifluor using a CoverWell (Grace Biolabs) and viewed on a Provis AX70 microscope (Olympus).
Scanning electron microscopy
Samples were fixed in culture dishes with 4% gluteraldehyde (TAAB) for 3 hours, washed 3 × 5 minutes in 0.1 M sodium cacodylate buffer pH7.2, followed by 90 minutes 1% osmium tetroxide in water, 2 × 10 minutes water washes, then dehydrated in an ascending series of ethanols, followed by 100% ethanol overnight. They were next transferred to 100% acetone in critical point drying holders and critical point dried using carbon dioxide in a Polaron E3000 critical point dryer, mounted on stubs with rapid araldite and sputter coated with gold in a Polaron E5100 sputter coater. Specimens were examined and images recorded using a Philips SEM501B scanning electron microscope fitted with a Deben Pixie 3000 digital scan generator and imaging system and operated at 15 kV.
Interactions of Bioglass® with regrowing axons in vivo
Surgery and subsequent histology
All animals were handled ethically in accordance with the guidelines of King’s College Health Guide for the Care and Use of Laboratory Animals. The Home Office for Animal Care Committee approved the experimental procedures. Twenty four male Wistar rats were randomized into 4 groups (A–D) and deeply anaesthetized using Isoflourane by administered inhalation. Under a dissecting microscope, the left sciatic nerve was transected in mid-thigh and a 0.5 cm gap was created between the proximal and distal nerve stumps. In group A, a silastic conduit (S.F. Medical, Hudson, MA) (internal diameter 1.98 mm, external diameter 3.17 mm, length 8 mm) containing an average of 10 Bioglass® fibres (0.5 cm long; ∼25 μm diameter) was sutured between the stumps using 10/0 sutures (Ethicon). Groups B–D were control groups. In group B, an empty silastic conduit was sutured between the nerve stumps, in group C a 0.5 cm section of the sciatic nerve was removed, reversed and sutured in place (i.e., an autograft), and in group D a piece of nerve was excised to leave a 0.5 cm inter-stump gap which was not treated further. In all cases, wounds were closed and animals were allowed to recover for 2 or 4 weeks. Animals were sacrificed at these times by CO2 asphyxiation (n = 3), and 3 cm lengths of nerve containing either the constructs (A, B), the autograft (C), or the tissue bridge that crossed the initially empty inter-stump gap (D), were harvested under a dissecting microscope and immersion-fixed overnight in 4% PFA. After fixation, and under a dissecting microscope, each silastic conduit in group A was gently separated from its contents. Most of the glass fibres were removed from the tissue thus exposed using fine forceps and the minimum of force: where resistance was encountered they were left in situ. Extracted glass fibres were stained with toluidine blue to assess the nature of any adherent material.
Tissue processing and immunohistochemistry
Processed tissue was embedded in polyester wax (Steedman, 1957). Each nerve was blocked out in a proximodistal sequence in order to sample the proximal and distal stumps and their associated outgrowth zones, and the mid inter-stump gap. Sections cut 1 cm distal to the proximal end of the distal stump were used for quantitative analysis of the extent of axonal regeneration. Transverse sections if 10 μm were immunostained with the following antibodies: anti-EHS laminin (1:200, Sigma), anti-S-100 (1:200, DAKO), anti-200kD neuro-filament (1:400, Serotec), and ED1 (1:200, Serotec) to reveal basal laminae, Schwann cells, axons and recruited macrophages, respectively. Primary antibodies were visualized using anti-mouse FITC or anti-rabbit TRITC (both Sigma 1:100). All immunostained samples were mounted in Citifluor and viewed on a Provis AX70 microscope (Olympus). Images were captured using an Axiocam H100 and Axiovision software (Carl Zeiss).
Quantification
S-100 positive Schwann cell tubes and neurofilament positive axons were counted in double immunolabelled transverse sections of each distal stump 1 cm distal to its proximal end, 4 weeks after surgery. The number of reinnervated Schwann cell tubes was expressed as a percentage of the total number of Schwann cell tubes in each section. Regeneration in group A was compared statistically to that seen in groups B–D using one-way ANOVA and Bonferroni’s multiple comparison post-tests (Prism 3.0, Graph Pad software).
RESULTS
Interactions of Bioglass® with Schwann cells in vitro
Primary cell cultures obtained from digested rat sciatic nerves contained ∼50% Thy-1 immunoreactive fibroblasts and ∼50% S-100 immunoreactive Schwann cells (data not shown). Phase microscopy and SEM revealed that bipolar spindle-shaped cells with a typical Schwann cell morphology and large fibroblast-like flattened cells had attached to both the fibres and the coverslips 24 hours after plating cells onto Bioglass® fibres or poly-
Interactions of Bioglass® with regrowing axons in vivo
In groups A, B and D, a tissue bridge extended between the proximal and distal stumps: the bridge was markedly thicker at 4 weeks than at 2 weeks in all animals, and was narrowest at both time points in group D. Regenerating axons grew within preexisting Schwann cell tubes in the autografted nerve (group C), and in minifasicles in groups A, B and D. (A minifascicle typically contains a small group of <20 myelinated and unmyelinated regrowing axons and their associated Schwann cells, the whole surrounded by one or two layers of perineurial cells: such structures may conveniently be identified by immunostaining the basal laminae delineating the Schwann cell tubes and the perineurial cells.) In group A, the minifascicles grew in well-vascularized connective tissue that was full of “holes” of varying sizes (Figs 1c and d). Presumably the holes had housed the Bioglass® fibres prior to their extraction at tissue harvesting. Although most holes appeared empty, some contained what appeared to be flakes of autofluorescent material, which we assume represented an early stage in the dissolution of the Bioglass®, and which may possibly correlate with the tesselated appearance of the fibres seen in SEM (Fig 1a). It was notable that this material did not cause sectioning artefacts, suggesting that it was softer than the preimplantation glass fibres, because this could not be sectioned satisfactorily. We suggest that these persisting fibres were the ones that had proved resistant to manual extraction. The connective tissue surrounding the holes contained occasional isolated, randomly oriented, S-110 positive Schwann cells: these cells were not usually associated with axons. Extraction of the glass fibres at harvesting appeared to cause minimal disruption to the surrounding tissue as assessed on tissue sections (Fig 1d): only small proteinaceous fragments were observed on the surface of extracted rods that had been stained with toluidine blue.
ED1 positive macrophages were present in the epineurium (groups A–D), the interstump tissue bridges (groups A, B, D), and the endoneurium of the autografts (group C).
Quantification of axonal regrowth through constructs
Axonal outgrowth from the proximal nerve stumps and the degree of penetration of the distal nerve stumps by regenerating axons was significantly more robust at both 2 and 4 weeks in groups A and C than in groups B and D (Fig 2). Statistical analysis revealed there was no significant difference in re-innervation between groups A and C (P>0.05), but that re-innervation was significantly greater in group A than in either of groups B or D (0.05 > P > 0.005 in both cases).
DISCUSSION
We have demonstrated in this proof of principle study that fibres of Bioglass® ensheathed within a silastic tube can support axonal regrowth across a 0.5 cm gap in an adult rat sciatic nerve. We used a well-established test system because we wished to examine the behaviour of the Bioglass® fibers and not the conduit. Quantitatively, the axonal regrowth that occurred in 4 weeks was statistically indistinguishable from that seen using an autograft and was at least ten times greater than the regrowth achieved in this time across either an empty silastic tube or an unrepaired gap.
Our results are consistent with previous findings that longitudinally aligned substrates facilitate axonal regeneration, whether natural, e.g. tendon (Brandt et al., 1999) or acellular sarcolemmal tubes (Enver and Hall, 1994; Hall, 1997), or artificially constructed, e.g. magnetically aligned type I collagen gel (Ceballos et al., 1999), or bundles of type I collagen filaments (Yoshii and Oka, 2001). On the basis of extrapolations from in vitro results (Dubey et al., 1999), the underlying mechanism is believed to be contact guidance of outgrowing axons and/or their associated non-neuronal cells.
Bioglass® has been shown to support the growth and differentiation of osteoblasts and osteoblast-like cells in vitro (Kaufmann et al., 2000; Loty et al., 2001; Lu et al., 2003). We have shown here for the first time that Bioglass® fibres support the attachment and spreading of Schwann cells in vitro: we believe that these fibres also provided a scaffold that was invaded by cells growing out from proximal and distal nerve stumps in vivo (Hall, 2004; Weis et al., 1994; Whitworth et al., 1995; Williams, 1987), and that these cells subsequently secreted an extracellular matrix that facilitated axonal regrowth.
That rodent peripheral axons can cross an interstump gap of ∼0.5 cm is well documented: the phenomenon has been reported in a variety of experimental models since the introduction of entubulation as a research tool over 20 years ago. However, we believe that our results are noteworthy for several reasons. Axonal regrowth was comparable with that seen using an autograft, suggesting that the lag time before axons and Schwann cells started to grow out from the proximal stumps was minimal and reinforcing the biocompatibility of the Bioglass® fibres. We noted that some of the Bioglass® fibres were already beginning to break down, which is an important consideration given that biocompatibility and bioresorbability are absolute requirements of successful tissue-engineered devices. We tested the fibres within a non-resorbable tube, the properties of which are well established. An optimized device would use a tube that elicited minimal tissue response in the wound bed and was bioresorbable, such as a biodegradable controlled release inorganic polymer glass (Gilchrist et al., 1998; Lenihan et al., 1998).
Recent evidence suggests that Bioglass® composites with a biodegradable polymeric phase may be useful as scaffolds in bone and lung tissue engineering (Verrier et al., 2004). In the context of nerve repair, bioengineered Bioglass® fibres could be used to deliver growth factors and/or adherent cells within the lumen of a nerve conduit. From a clinical perspective, the fibres can be many centimetres long, packed into conduits of varying internal diameters, and their manufacture is independent of the time of use: Bioglass® fibres therefore have the potential to be used in the most challenging cases, where inter-stump gaps are longer and nerves are wider than those used in this study.
Footnotes
Acknowledgements
We are grateful to Dr Tony Brain, Centre for Ultrastructural Imaging, King’s College London, for expert technical assistance with SEM and to Garrit Koller, GKT Dental Institute, for preparation of the Bioglass® fibres.