Peripheral nerve regeneration with sustained release of poly(phosphoester) microencapsulated nerve growth factor within nerve guide conduits

Biomaterials 24 (2003) 2405–2412

Peripheral nerve regeneration with sustained release of poly(phosphoester) microencapsulated nerve growth factor within nerve guide conduits Xiaoyun Xua, Woon-Chee Yeeb, Peter Y.K. Hwangb, Hanry Yua, Andrew C.A. Wana, Shujun Gaoa, Kum-Loong Boonb, Hai-Quan Maoc, Kam W. Leonga,c, Shu Wanga,d,* a

Molecular and Biomaterials Lab., Institute of Materials Research and Engineering, National University of Singapore, 3 Research Link, Singapore 117602, Singapore b National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore c Department of Biomedical Engineering, Johns Hopkins University, Baltimore 21205-2196, USA d Department of Biological Science, National University of Singapore, Singapore 117597, Singapore Received 29 October 2002; accepted 19 January 2003

Abstract Prolonged delivery of neurotrophic proteins to the target tissue is valuable in the treatment of various disorders of the nervous system. We have tested in this study whether sustained release of nerve growth factor (NGF) within nerve guide conduits (NGCs), a device used to repair injured nerves, would augment peripheral nerve regeneration. NGF-containing polymeric microspheres fabricated from a biodegradable poly(phosphoester) (PPE) polymer were loaded into silicone or PPE conduits to provide for prolonged, site-specific delivery of NGF. The conduits were used to bridge a 10 mm gap in a rat sciatic nerve model. Three months after implantation, morphological analysis revealed higher values of fiber diameter, fiber population and fiber density and lower Gratio at the distal end of regenerated nerve cables collected from NGF microsphere-loaded silicone conduits, as compared with those from control conduits loaded with either saline alone, BSA microspheres, or NGF protein without microencapsulation. Beneficial effects on fiber diameter, G-ratio and fiber density were also observed in the permeable PPE NGCs. Thus, the results confirm a longterm promoting effect of exogenous NGF on morphological regeneration of peripheral nerves. The tissue-engineering approach reported in this study of incorporation of a microsphere protein release system into NGCs holds potential for improved functional recovery in patients whose injured nerves are reconstructed by entubulation. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Nerve growth factor; Polymeric microsphere; Nerve guide conduits; Poly(phosphoester); Nerve regeneration

1. Introduction Synthetic nerve guide conduits (NGCs) have been widely tested in pre-clinical studies to repair nerve defects [1–3]. This approach circumvents problems inherited with the current clinical treatment of using autologous nerve grafts, such as functional loss at the donor sites, formation of potential painful neuromas, structural differences between donor and recipient nerves and shortage of graft materials in the case of extensive repair. However, the synthetic conduits did not function as well as nerve autografts in most cases. *Corresponding author. Tel.: +65-874-8530; fax: +65-872-7150. E-mail address: [email protected] (S. Wang).

Several tissue-engineering approaches have been proposed to enhance the performance of conduits, which include delivering neurotrophic factors within hollow tubes. Nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), fibroblast growth factor, glial growth factor and ciliary neurotrophic factor delivered within a conduit may significantly increase the morphological and/or functional recovery of transected and repaired nerves [4–13]. It was, however, noted that while delivery of NGF, a neurotrophic protein predominantly acting on sensory and sympathetic neurons, promoted nerve regeneration within conduits at an early stage, the promoting effect would not last after 1 month [6,9,13]. This was probably due to the rapid decline of NGF concentrations in the conduit

0142-9612/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0142-9612(03)00109-1

2406

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

caused by the degradation in aqueous media at 37
C, leakage from the conduit and/or dilution by entering fluids. Thus, instead of applying naked proteins in NGCs during implantation, use of a configuration that may provide sustained protein release as well as protein protection against degradation would be more appealing. Polymeric microspheres have been developed for the encapsulation of various kinds of neurotrophic factors and for local sustained delivery within the nervous system [14–23]. We have developed a sustained NGF release system using a biodegradable poly(phosphoester) (PPE) polymer [24]. Bioactive NGF could be released from these microspheres for at least 10 weeks, as demonstrated by neurite outgrowth assay of the PC12 cells [24]. In this study, we tested whether incorporation of this sustained NGF release system into NGCs would improve the regeneration of peripheral nerves.

into the non-solvent bath where it was allowed to stand for 5 min. The mandrel was subsequently rotated horizontally for 5 min to reduce variations in the wall thickness along the axis of the tube and at the same time, to facilitate the process of air-drying. Three coating steps were used to obtain a polymer tube with wall thickness of about 170 mm. The coated mandrel was subjected to either freeze-drying or direct vacuum drying. For freeze-drying, the coated mandrels were equilibrated in water overnight, frozen at 20
C and subsequently freeze-dried using a Modulyo freezedrying Unit at a pressure of 0.1 Torr for at least 1 week. For direct vacuum drying, a similar vacuum and period of drying was employed without freezing the sample. Finally, the polymer coatings were removed from the mandrel and cut to 14 mm lengths for implantation.

2.3. NGC implantation 2. Materials and methods 2.1. Fabrication of NGF microspheres NGF microspheres were prepared by a modified W/ O/W emulsion solvent evaporation/extraction method with PPE, P(DAPG-EOP) and 2.5S NGF (Life Technologies, purified from submaxillary gland of male mice) as published previously [24,25] In brief, the method involved the use of three phases: (1) an inner water phase (W1) containing the protein to be incorporated, (2) an intermediate organic phase consisting of a polymer/methylene chloride solution, and (3) an outer water phase (W2) containing an emulsifying agent. About 1 mg NGF, 30 ml of BSA solution (100 mg/ml) and 300 ml of P(DAPG-EOP) polymer solution (5%, w/v) were used. BSA was used as a carrier protein to modulate the release rate of NGF. The loading level achieved for NGF was 0.003% (w/w). The size of P(DAPG-EOP) microspheres is from 12.2+4.4 to 15.3+6.2 mm. Microspheres containing FITC-labeled BSA was also prepared to examine the distribution of the microspheres in regenerated tissues. 2.2. Fabrication of PPE nerve guide conduits (NGCs) Polymer synthesis and NGC fabrication were carried out according to the reported methods [24,25]. A 30% (w/w) solution of PPE polymer, P(BHET-EOP/TC) (MW. 19,000) in chloroform was prepared by magnetic stirring. A Teflon mandrel of diameter 1.5 mm was vertically dipped into the polymer solution by a mechanical linear head at a speed of 8.3 mm/s and allowed to immerse in the solution for 30 s. The mandrel was withdrawn at 24 mm/s and immediately immersed

Male adult Wistar rats weighing around 200 g were obtained from National University of Singapore Animal Center. The use of these rats in the current study was approved by the Animal Ethical Committee of National University of Singapore. The rats were kept on soft bedding in transparent cages at room temperature with a 1:1 light: dark cycle and free access to food and water. After anaesthetizing with pentobarbital, the right sciatic nerve of the rat was exposed through a 2 cm long skin incision on the thigh and retraction of the gluteus maximus muscle. The nerve was freed from surrounding tissue and transected at the midthigh level, proximal to the tibial and peroneal bifurcation. A 5– 7 mm piece of the nerve was removed and then the proximal and distal nerve stumps were pulled 2 mm into each opening of silicone or PPE nerve guide tubes, leaving a 10 mm interstump gap. The two stumps were fixed within the tubes with a single 10/0 perineurial suture (Ethilon). Before the proximal stump was pulled into the tube opening, the tube was filled with either saline, a saline solution with 50 ng NGF or a saline suspension of PPE microspheres containing 100 ng NGF. The surgery was performed under an Olympus operating microscope. The muscle layers were closed with 4/0 silk sutures and the skin closed with Michel clips. Silicone tubes made of medical-grade silicone tubing (Norton, OH, USA) possessed an internal diameter of 1.587 mm and external diameter of 3.175 mm. PPE tubes, fabricated according to the procedure in the last section, had an internal diameter of 1.5 mm and external diameter of 1.67 mm. The internal diameters of the tubes were chosen so as not to create any compression due to swelling of the enclosed nerve segment. All tubes used for nerve repair were 14 mm in length.

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

2407

2.4. Evaluation of nerve regeneration After 3 months, the rats were anaesthetized again and the sciatic nerves together with tubes were exposed and carefully isolated from the surrounding tissue. The nerve segment distal to the tube was pinched with a pair of forceps. Contraction of muscles on the back or retraction of the leg indicated the presence of regenerating sensory fibers in the pinching segment, while no response was taken as an indication of the absence of such fibers. For histological examination, the rats were sacrificed with an overdose of pentobarbital. The nerve guide tube was opened by a longitudinal incision and its contents photographed. The tube contents, together with a 3–5 mm nerve segment distal to the tube, were then fixed into 2.5% glutaradelyde in PBS buffer (pH=7.4) for at least 3 days. The subsequent fixation, embedding, sectioning, and staining procedures were the same as previously described [26]. Quantitative evaluation was carried out at the distal segment of the regenerated nerve using Micro Image Litet (Olympus, Image Analysis Software). Areas of interest were selected for ultrathin sectioning. The ultrathin sections (100 nm) were stained with lead citrate, collected on copper mesh grids and examined in Philips EM 208s electron microscope operating at 80–100 kV. The samples were evaluated for regeneration of nerve tissue and foreign body reaction against the polymers.

3. Results As microsphere distribution in a regenerated tissue cable may affect local protein concentrations, we first investigated the distribution using FITC-BSA-loaded PPE microspheres (Fig. 1A) within the silicone NGC, a type of conduit that is not degradable and impervious. Under the confocal microscope, the microspheres diffused through the fibrin cable (Fig. 1B) and did not precipitate on one side of the NGC. Three months after implantation of silicone NGCs, reflex responses were examined by pinching regenerated nerve trunks distal to the conduits in anesthetized animals. The positive reflex response was observed in 36% (5 out of 14) of the rats in the saline control group (Group I), 60% (6/10) in the BSA microsphere control group (Group II), 45% (9/20) in the naked NGF control group (Group III), and 62% (8/13) of the rats with NGF-containing microspheres-loaded conduits (Group IV). All these rats had a regenerated cable inside the conduits, which had bridged a 10-mm gap between the nerve stumps. The regenerated nerve cables were centrally located within the conduits, surrounded by a fine epineurium. The cables contained numerous fascicles of myelinated as well as unmyelinated axons (UMAs). Most of the axons in the distal nerve trunks

Fig. 1. Distribution of microspheres within a regenerated tissue cable. Three milligrams of FITC-BSA was used for microsphere preparation. Protein:polymer ratio is 1:5 and PVA concentration is 10%. (A) FITCBSA distribution in P(DAPG-EOP) microspheres as showed through confocal scanning pictures. (B) Distribution of FITC-BSA/P(DAPGEOP) microspheres in regenerated tissues inside in an NGC. The sample was collected 7 days after conduit implantation.

were already myelinated. Transverse sections at the distal part of the 10-mm gaps were analyzed to determine the number of regenerated axons, fiber density, and fiber diameters (Fig. 2). There were no significant value differences in all these parameters among the control groups. The samples collected from NGF microspheres-loaded NGCs had more myelinated axons (MA), higher fiber density and thicker myelin sheath than those from the control groups (Table 1). We also used NGCs made of a biodegradable PPE polymer, P(BHBT-EOP/TC), to test effects of NGF microspheres. Three months after implantation, when the nerve trunks distal to the conduits were pinched in anesthetized animals, positive reflex response were observed in 40% (4/10) of the rats in the saline control group (Group V); 50% in the BSA microsphere control group (Group VI) and 70% (7/10) of the rats

2408

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

in the PPE plus NGF encapsulated microsphere group, as compared to the control groups (Fig. 3).

4. Discussion Previous studies of the rat sciatic nerve model have demonstrated an increase in nerve regeneration at 3 or 4 weeks following introduction of exogenous NGF in a free protein form into silicon NGCs [6,9,13]. After that period, no improvement in overall histological appearance could be seen [6, 9], probably because of the failure in maintaining sufficient NGF concentrations over the long duration of nerve regeneration. Other studies using various nerve conduits in combination with single administration of NGF in different animal models showed basically the same finding that NGF may enhance sensory [4,12] and motor [7] nerve regeneration between 3 and 7 weeks. The major finding of the present study is the improved nerve regeneration for at least 3 months after applying a new tissue-engineering approach incorporating a microsphere NGF delivery system into NGCs. Understanding the features associated with our method and its application may help to develop more effective therapeutic approaches that promote functional recovery of injured nerves. 4.1. Poly(phosphoester) for NGF delivery

Fig. 2. Effects of NGF-loaded microspheres on nerve regeneration with silicon tubes. Micrographs of semithin cross sections of distal nerve segments from rats implanted with a conduit loaded with BSA (A) and with NGF/BSA (B), respectively. Note the increased number of MAs in (B); Toluidine blue staining. The original magnification  1000.

with NGF-loaded microspheres (Group VII). All these rats possessed a regenerated tissue cable inside the conduits that had bridged a 10-mm gap. Morphological analysis was carried out in transverse sections through the distal part of the 10-mm gaps. Similar to what was observed in the silicon NGCs, improvements in nerve regeneration were observed after applying NGF microspheres in PPE NGC, with significant increases in fiber diameter and fiber population and decrease in G-ratio (Table 1). The electron microscopic evaluation showed results consistent with light microscopic results: there was a higher number of larger axons with thicker myelin

Poly (lactide-co-glycolide)s (PLGAs) are probably the most popular and well-characterized biodegradable polymeric biomaterials. The regulatory approval and extensive database for their human use render them an obvious choice for medical applications that range from controlled drug delivery to tissue engineering [27,28]. Several NGF microsphere delivery systems made of PLGA had previously been developed [14,15,17,22,23]. The protein release profiles of these systems were often characterized by a marked initial burst followed by a slow continuous release [15,22]. The subsequent release of NGF could be low in relation to the initial burst [14]. Inactivation of the protein by the acidic degradation products of the polymer is another concern [29]. To improve upon these shortcomings, we investigated the use of polyphosphoesters (PPEs) for the delivery of this growth factor. PPE is a class of biodegradableble polymers with a variety of attractive properties [25,30,31]. The polymers are adjustable; manipulation of either the backbone or the sidechain structure would readily alter their physicochemical properties. The phosphoester bond in the PPE backbone can be cleaved by water under physiological conditions. The polymers are also characterized by low inflammatory responses and no toxic effects on neurons [26]. We have successfully used one PPE polymer, P(BHET-EOP/TC), as an NGC material [26].

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

2409

Table 1 Morphometric analysis of the regenerated nerves at the distal end of conduits 3 months after implantation A: Mean7SEM Group: conduits

Treatment

N

Fiber diameter (mm)

G-ratio

Fiber population

Fiber density (mm2)

I: Silicon II: Silicon III: Silicon IV: Silicon V: PPE VI: PPE VII: PPE

Saline BSA microspheres NGF protein NGF microspheres Saline BSA microspheres NGF microspheres

5 6 9 8 4 5 7

3.3670.14 3.2470.27 3.4770.21 3.7970.29 3.8570.20 3.0270.24 4.1470.28

0.7870.03 0.7770.50 0.8170.03 0.7470.03 0.8170.02 0.7970.01 0.7670.03

502472317 492171257 596772669 1083072906 596772389 857373762 1050073230

1718273273 1053078777 1701773626 2171273473 1244472202 1225272468 1910473681

B: Turkey test for multiple comparisons in groups with significant ANOVA Group

Fiber diameter

G-ratio

Fiber population

Fiber density

I vs. III I vs. II I vs. IV II vs. IV III vs. IV V vs. VI V vs. VII VI vs. VII

NS NS po0:01 Po0:01 po0:05 po0:01 NS po0:01

NS NS po0:05 NS po0:001 NS po0:05 po0:05

NS NS po0:01 po0:01 po0:01 NS p ¼ 0:01 NS

NS NS po0:05 po0:01 p ¼ 0:05 NS po0:05 po0:05

P(DAPG-EOP) is another such polymer, having phosphate bonds distributed between oligomeric blocks of lactides in the backbone. The rationale for designing such a structure is to explore the possibility of extending the physicochemical properties of the PLGA polymers. The inclusion of the phosphate linkage in the backbone offers an extra degree of freedom to fine-tune the properties of the polymer [25]. The degradation rate of this kind of polymers is mainly controlled by the percentage of the phosphate component introduced into the backbone. The higher the phosphate content of the backbone, the faster the degradation rate of the polymer. An additional factor that determines the overall degradation rate is the stereoisomerism of the lactide, with the d,l-lactide producing polymers with low crystallinity, low Tg and fast degradation rate. Similar to many other polypeptides, NGF does not cross the blood–brain barrier. Its application requires direct administration to the brain, and hence can benefit significantly from the controlled release approach. NGF has a limited stability under physiological conditions [32]. Its stability may be further reduced in the case of protein encapsulation by exposure to organic solvents, shearing, and acidic degradation products. We have demonstrated the bioactivity of NGF released from PPE microspheres over a 10-week period in PC12 cell cultures [24]. The long stability of encapsulated NGF may result from a combination of factors. For example, co-encapsulation of BSA may stabilize NGF by providing some buffering capacity [15]. PPE polymers have

been shown to degrade by a combined mechanism of surface erosion and bulk degradation [26], compared to predominantly bulk degradation of PLGA polymers that will lower the pH of the polymer matrix considerably. Surface erosion of PPE polymers also suggests less interaction between polymer degradation products and encapsulated proteins. The polymer P(DAPG-EOP) contains phosphate bonds distributed between oligomeric blocks of lactides in the backbone, likely producing fewer lactides than PLGA when it is broken down. In view of the high sensitivity of proteins to acidic degradation products, P(DAPG-EOP) therefore offers the advantage of a biodegradable material for fabrication of microspheres that may maintain a relatively high bioactivity of encapsulated proteins for a long period of time. 4.2. NGF effects within NGCs NGF was first recognized for its stimulatory effect on nerve fiber growth, probably via an action on transcription-dependent processes in the neuronal cell body [33– 35]. Later experiments have shown the neurite extension requires direct action by NGF on the growth cones, the elongating machinery at the end of a growth nerve process [36]. Moreover, local sources of NGF can dictate the direction of this neuritic growth [37]. NGF released from our PPE microspheres may act at several sites in an NGC. In adult rats, sciatic nerve transection causes cell death in a portion of the neuronal

2410

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

Fig. 3. EM morphology of distal nerve segment following nerve section and grafting with PPE conduit (A), and grafting with PPE plus NGF (B) impregnated microspheres (Day 90 post graft). Both MA and UMA are associated with SC. Possible neurite sprouting (NS) in a sample collected from an NGF microspheres-loaded NGC is shown in (C). Magnification:  6300, Bar=0.1 mm.

population in the dorsal root ganglia (DRG) or in the ventral horn of the spinal cord, with 20–40% of DRG neurons being lost [38]. The cell loss is associated with atrophy in the proximal nerve stump [39] and may be counteracted by NGF administered in silicone tubes

fixed to the proximal nerve stump [11]. NGF internalization and transport is mediated by high-affinity Trk A NGF receptors [38]. The expression of these receptors in adult sensory neurons may be increased by NGF [40] and depleted by axotomy [41]. It is also shown that a crush lesion of the sciatic nerve induced increases in receptor-mediated retrograde transport of 125I-labeled NGF that was injected into the lesion site [42]. This increase, together with a continued supply of NGF such as that provided by NGF-loaded microspheres in the present study to the proximal stump of a transected nerve, would maintain or increase expression of highaffinity NGF receptors on neurons. A potential outcome of increased expression of NGF receptors would be a further increase in the retrograde transport of NGF to the neurons and subsequent neuronal survival. NGF in DRG could strongly stimulate gene expression of BDNF [43] and thus protect BDNF-response neurons. These remote actions after NGF transport may maintain neurofilament contents and axonal calibers after axonal injury [44]. They may also serve as one mechanism underlying the improvement in fiber diameter after applying microencapsulated NGF in the present study. NGF may act locally on regenerating nerve fibers within NGCs. This protein is well documented for its chemoattractive effect on developing and regenerating axons. Its concentration gradients may lead to an oriented outgrowth of axons to their targets [36] and administration of NGF antiserum prevents this outgrowth [45]. During regeneration of the peripheral nervous system, diffusible concentration gradients of tropic factors released from denervated nerves may direct the outgrowth of adult sensory and motor axons over distances of more than 6.5 mm [46]. The spatial– temporal analysis of the progress of nerve regeneration across a 10-mm rat sciatic nerve gap repaired with a silicone tube has shown that, without any other help, it would take 2 weeks for growth cones of regenerating axons to migrate into the gap between the two nerve stumps and 4 weeks for them to reach the distal end [3]. If the chemoattractive effect of NGF is critical for neurite outgrowth, a continued supply of exogenous NGF within NGCs by a release system like that provided by microspheres would be needed. A single administration of high doses of NGF at mg/ml levels into NGCs during tube implantation may maintain physiologically relevant amounts of NGF in the chamber fluids for about 2 weeks and enhance the initial outgrowth of neuronal fibers into the chamber [6]. This is consistent with what we observed in a previous study using NGF-loaded conduits where NF68-positive fibers were detected at the distal stump 2 week after implantation [24]. However, with the degradation and inactivation of NGF, a single dose administration of the protein was of no help to nerve regeneration at later

X. Xu et al. / Biomaterials 24 (2003) 2405–2412

periods as demonstrated by others [6] and us in the present study. NGF may also act locally on Schwann cells (SC). An improved G-ratio in the microencapsulated NGFtreated group noted in our study supports the notion that sustained delivery of NGF may benefit the myelinizing process induced by SC [47,48]. NGF appears to have no direct influence on mitosis of SC, but does promote the migration of these cells through its low-affinity receptors on the SC surface [47]. It is of note that following transection in mature peripheral nerves, SC in the distal segment up-regulate the expression of their low affinity receptors by at least 50-fold [48]. Exogenous NGF binding to these receptors may, therefore, facilitate the migration of SC into the gap between the injured stumps and formation of cords along which growth cones of regenerating axons may migrate toward the distal end. Immobilized and concentrated NGF proteins on the SC surface may also exhibit a chemotactic effect or serve as a permissive substratum when a regenerating axon approaches. Upon completion of nerve regeneration, the conduit structure no longer serves any purpose. It may, in fact, become detrimental due to mechanical impingement or infection. Biodegradable NGCs potentially avoid these problems, besides making secondary surgery unnecessary. In the previous study, PPE tubes have been proven to be effective in nerve reconstruction [26]. But when combined with NGF and NGF-loaded microspheres, more parameters had to be taken into consideration. Tube integrity was one of the key issues. In PPE tubes, some cracks started to appear as early as 1–3 days after implantation [26]. The tube porosity [49] and increased permeability due to tube cracks had allowed the influx of nutrients and growth factors from the surrounding environment and enhanced the constitution of the matrix. These could have resulted in leakage of preloaded NGF and lower its concentrations within the conduits, whereas the enclosed environment of silicone tubes would maintain local concentrations of the growth factors [50,51]. This might be the reason why some parameters of nerve regeneration with PPE tubes were not improved significantly. Our observation is to some extent consistent with the finding reported by Derby et al. [6] that with NGF treatment, impermeable tubes were more effective in promoting regeneration of neuronal fibers than macroporous semipermeable tubes.

5. Conclusions We have shown that sustained release of NGF for a prolonged period from microspheres loaded into NGCs may improve peripheral nerve regeneration. The advantages of the novel application of microspheres include sustained local action of a trophic factor and

2411

reduction of the total amount of trophic factor proteins needed to generate a satisfactory biological effect. The approach potentially offers the flexibility of incorporation into one NGC of multiple delivery systems for different neurotrophins that would be released with different profiles and act on the different stages or cellular components of nerve regeneration.

Acknowledgements This work was supported by funding from the National Science and Technology Board of Singapore to Institute of Materials Research and Engineering and National Medical Research Council of Singapore to National Neuroscience Institute.

References [1] Dahlin LB, Zhao Q, Bjursten LM. Nerve regeneration in silicone tubes: distribution of macrophages and interleukin-1b in the formed fibrin matrix. Rest Neurol Neurosci 1995;8:199–203. [2] Danielsen N, Varon S. Characterization of neurotrophic activity in the silicone-chamber model for nerve regeneration. J Reconstr Microsurg 1995;11:231–5. [3] Valentini RF, Aebisher P. Strategies for the engineering of peripheral nervous tissue regeneration. In: Lanza RP, Langer R, Chick WL, editors. Principles of tissue engineering. Austin, TX, USA: R.G. Landes Company, 1997. p. 671–84. [4] Chen YS, Wang-Bennett LT, Coker NJ. Facial nerve regeneration in the silicone chamber: the influence of nerve growth factor. Exp Neurol 1989;103:52–60. [5] Bryan DJ, Holway AH, Wang KK, Silva AE, Trantolo DJ, Wise D, Summerhayes IC. Influence of glial growth factor and Schwann cells in a bioresorbable guidance channel on peripheral nerve regeneration. Tissue Eng 2000;6:129–38. [6] Derby A, Wayne EV, Frierdich GE, Neises G, Rapp SR, Roua DG. Nerve growth factor facilitates regeneration across nerve gaps: morphological and behavioral studies in rat sciatic nerve. Exp Neurol 1993;119:176–91. [7] He C, Chen Z. Enhancement of motor nerve regeneration by nerve growth factor. Microsurgery 1992;13:151–4. [8] Ho PR, Coan GM, Cheng ET, Niell C, Tarn DM, Zhou H, Sierra D, Terris DJ. Repair with collagen tubules linked with brainderived neurotrophic factor and ciliary neurotrophic factor in a rat sciatic nerve injury model. Arch Otolaryngol Head Neck Surg 1998;124:761–6. [9] Hollowell JP, Villadiego A, Rich KM. Sciatic Nerve regeneration across gaps within silicone chambers: long-term effects of NGF and consideration of axonal branching. Exp Neurol 1990;110: 45–51. [10] Laquerriere A, Peulve P, Jin O, Tiollier J, Tardy M, Vaudry H, Hemet J, Tadie M. Effect of basic fibroblast growth factor and amelanocytic stimulating hormone on nerve regeneration through a collagen channel. Microsurgery 1994;15:203–10. [11] Otto D, Unsicker K, Grothe C. Pharmacological effects of nerve growth factor and fibroblast growth factor applied to the transectioned sciatic nerve on neuron death in adult rat dorsal root ganglia. Neurosci Lett 1987;83:156–60. [12] Pu LLQ, Syed SA, Reid M, Patwa H, Goldstein JM, Forman DL, Thomson JG. Effects of nerve growth factor on nerve

2412

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27] [28]

[29]

[30]

X. Xu et al. / Biomaterials 24 (2003) 2405–2412 regeneration through a vein graft across a gap. Plast Reconstr Surg 1999;104:1379–85. Rich KM, Alexander TD, Pryor JC, Hollowell JP. Nerve growth factor enhances regeneration through silicone chamber. Exp Neurol 1989;105:162–70. Camarata PJ, Suryanarayanan R, Turner DA, Parker RG, Ebner TJ. Sustained release of nerve growth factor from biodegradable polymer microspheres. Neurosurgery 1992;30:313–9. Cao XD, Shoichet MS. Delivering neuroactive molecules from biodegradable microspheres for application in central nervous system disorders. Biomaterials 1999;20:329–39. Krewson CE, Dause R, Mak M, Saltzman WN. Stabilization nerve growth factor in controlled release polymers and in tissue. J Biomater Sci Polym Ed 1996;8:103–17. Lamm XM, Duenas ET, Cleland JL. Encapsulation and stabilization of NGF into PLGA microspheres. J Pharm Sci 2001;90:1356–64. Menei P, Benoit JP, Boisdron-clle M, Fournier D, Mercier P, Guy G. Drug targeting into the central nervous system by stereotactic implantation of biodegradable microspheres. Neurosurgery 1994;34:1058–64. Menei P, Daniel V, Montero-Menei C, Brouillard M, Barthelaix BP, Benoit JP. Biodegradation and brain tissue reaction to poly(d,l-lactide-co-glycolide) microspheres. Biomaterials 1993;14:470–8. Menei P, Benoit P, Boisdronclle M, Croue A, Guy G. Effect of stereotactic implantation of biodegradable 5-fluorouracil-loaded microspheres in healthy and C6 glioma bearing rats. Neurosurgery 1996;39:117–23. Mital S, Cohen A, Maysinger D. In vitro effects of brain derived neurotrophic factor released from microspheres. Neuroreport 1994;5:2577–82. Pean JM, Venier-Julienne MC, Boury F, Menei P, Denizot B, Benoit JP. NGF release from poly(d,l-lactide-co-glycolide) microspheres: effect of some formulation parameters on encapsulated NGF stability. J Control Rel 1998;56:175–87. Saltzman WM, Mak MW, Mahoney MJ, Duenas ET, Cleland JL. Intracranial delivery of recombinant nerve growth factor: release kinetics and protein distribution for three delivery systems. Pharm Res 1999;16:232–40. Xu XY, Yu H, Mao HQ, Gao SJ, Leong KW, Wang S. Polyphosphoester microspheres for sustained release of biologically active nerve growth factor. Biomaterials 2002;23:3765–72. Mao H-Q, Shipanova-Kadiyaia I, Zhao Z, Dang W, Leong KW. Encyclopedia of controlled drug delivery. In: Mathiowitz E, editor. Biodegradable polymers: poly (phosphoester)s. New York: Wiley, 1999. p. 45 [chapter III]. Wang S, Wan ACA, Xu XY, Gao SJ, Mao HQ, Leong KW, Yu H. A new nerve guide conduit material composed of a biodegradable poly(phosphoester). Biomaterials 2001;22: 1157–69. Cleland JL. Protein delivery from biodegradable microspheres. Pharm Biotechnol 1997;10:1–43. Shive MS, Anderson JM. Biodegradation and biocompatibility of PLA and PLGA microspheres. Adv Drug Delivery Rev 1997;28:5–24. Zhu G, Mallery SR, Schwendeman SP. Stabilization of proteins encapsulated in injectable poly (lactide-co-glycolide). Nat Biotechnol 2000;18:52–7. Dahiyat BI, Hostin E, Posadas EM, Leong KW. Synthesis and characterization of putrescine-based poly (phosphoester-urethanes). J Biomater Sci Polym Ed 1993;4:529–43.

[31] Richards M, Dahiyat BI, Arm DM, Brown PR, Leong KW. Evaluation of polyphosphates and polyphosphonates as degradable biomaterial. J Biomed Mater Res 1991;25:1151–67. [32] Server AC, Shooter EM. Nerve growth factor. Adv Protein Chem 1977;31:339–409. [33] Ebendal T, Soderstrom S, Hallbook F, Ernfors P, Ibanez CF, Persson H, Wetmore C, Stromberg I, Olson L. Human nerve growth factor: biological and immunological activities, and clinical possibilities in neurodegenerative disease. Adv Exp Med Biol 1991;269:207–25. [34] Ebendal T. Function and evolution in the NGF family and its receptors. J Neurosci Res 1992;32:461–70. [35] Hellweg R, Raivich G. Nerve growth factor: pathophysiological and therapeutic implications. Pharmacopsychiatry 1994;27:15–27. [36] Campenot RB. Local control of neurite sprouting in cultured sympathetic neurons by nerve growth factor. Brain Res 1987;45:293–301. [37] Gundersen RW, Barrett JN. Characterization of the turning response of dorsal root neuritis toward nerve growth factor. J Cell Biol 1980;87:546–54. [38] Terenghi G. Peripheral nerve regeneration and neurotrophic factors. J Anat 1999;194:1–14. [39] Schmalbruch H. Loss of sensory neurons after sciatic nerve section in the rat. Anat Res 1987;219:323–9. [40] Lindsay RM, Shooter EM, Radeke MJ, Misko TP, Dechant G, Thoenen H, Lindholm D. Nerve growth factor regulates expression of the nerve growth factor receptor gene in adult sensory neurons. Eur J Neurrosci 1990;2:389–96. [41] Verge VMK, Riopelle RJ, Richardson PN. Nerve growth factor receptors on normal and injured sensory neurons. J Neurosci 1989;9:916–22. [42] DiStefano PS, Curtis R. Receptor mediated retrograde axonal transport of neurotrophic factors is increased after peripheral nerve injury. Prog Brain Res 1994;103:35–42. [43] Apfel SC, Wright DE, Wiideman AM, Dormia C, Snider WD, Kessler JA. Nerve growth factor regulates the expression of brainderived neurotrophic factor mRNA in the peripheral nervous system. Mol Cell Neurosci 1996;7:134–42. [44] Gold BG, Mobley WC, Matheson SF. Regulation of axonal caliber, neurofilament content, and nuclear location in mature sensory neurons by nerve growth factor. J Neurosci 1991;11: 943–55. [45] Diamond J, Holmes M, Coughlin M. Endogenous NGF and nerve impulses regulate the collateral sprouting of sensory axons in the skin of the adult rat. J Neurosci 1992;12:1454–66. [46] Zheng M, Kuffler DP. Guidance of regenerating motor axons in vivo by gradients of diffusible peripheral nerve-derived factors. J Neurobiol 2000;42:212–9. [47] Anton ES, Weskamp G, Reichardt LF, Matthew WD. Nerve growth factor and its low-affinity receptor promote Schwann cell migration. Proc Natl Acad Sci USA 1994;91:2795–9. [48] Taniuchi M, Clark HB, Johnson Jr EM. Induction of nerve growth factor receptor in Schwann cells after axotomy. Proc Natl Acad Sci USA 1986;83:4094–8. [49] Wan ACA, Mao HQ, Wang S, Leong KW, Yu H. Fabrication of poly(phosphoester) nerve guides by immersion precipitation and the control of porosity. Biomaterials 2001;22:1147–56. [50] Longo FM, Hayman EG, Davis GE. Neurite-promoting factors and extracellular matrix components accumulating in vivo within nerve regeneration chambers. Brain Res 1984;309:105–17. [51] Lundborg G, Longo FM, Varon S. Nerve regeneration model and trophic factors in vivo. Brain Res 1982;232:157–61.