Author:Zhenwei Wang, Na Han, Jiancheng Wang, Hua Zheng, Jianping Peng, Yuhui Kou, Chungui Xu, Shuai An, Xiaofeng Yin, Peixun Zhang, Baoguo Jiang

Abstract:

Objective: To evaluate the long-term results of the use of nerve growth factor (NGF)-loaded poly-D, L-lactide-co-glycolide (PLGA) microspheres for improve nerve regeneration with small gap tubulization. Methods: NGF microspheres were prepared by a modified W/O/W emulsion solvent evaporation method. Forty-eight male SD rats were separated into 4 groups and received a chitin conduit to bridge a sciatic nerve injury left a 2 mm gap. Saline (Group A), 20 ng/ml NGF solution (Group B), blank PLGA microspheres (Group C), or 40 ng/ml NGF-loaded microspheres (Group D) was injected in the gap. Each group had two study endpoints, 3 months subgroup and 1 year subgroup. Results: The myelinated fiber count at 2 mm distal to the conduit at 1 year was slightly less than at 3 months in all groups (P>0.05). However, the maturity of the myelinated nerves at 1 year was obviously improved. The fiber count, myelin sheath thickness, axon area of NGF microsphere group were significantly higher than the saline groups at 3 months (P=0.05, P<0.05, and P<0.05, respectively). The SFI was significantly improved in NGF microspheres group compared to the saline group and NGF solution group at 1 year (P<0.05, and P<0.05, respectively). Conclusions: The results demonstrated that the release of NGF microspheres in small gap tubulization benefit on peripheral nerve injury facilitated nerve regeneration histologically, especially for the maturity of early regenerative nerve fibers and also had an effect on functional recovery in the long term.

Keywords: Nerve growth factor (NGF), microsphere, nerve regeneration, small gap, tubulization

Introduction

Repair of peripheralnerve injuries is an intrac­table problem in the clinic. Epineurial neuror­rhaphyhas long been performed as a tradition­al repair method, but functionalrecovery is typically unsatisfactory. In our previous study, we observed afavorable effect of small gap tubulization over the epineurial neurorrhaphyapproach. However, functional recovery to the pre-injury level remains limited.Thus, enhancing the effect of the conduits has become a focus in the field ofperipheral nerve regeneration.

The use of NGF for improving nerve regenera­tion has been welldocumented, but the obser­vation time of prior study was rarely beyond 3months. The purpose of this study was to evaluate the long term results of theuse of nerve growth factor for improve nerve regenera­tion with small gaptubulization. To solve the fast degradation and metabolism of NGF underphysiological conditions, NGF loaded poly-D, L-lactide-co-glycolide (PLGA)microspheres were used to release NGF slowly and protect the bioactivity.

Materials andmethods

Ethics statement

The study was approved by the Research Ethics Committee atPeking University People’s Hos-pital and met international biomedical ethics guidelines.The Biostatistics Department of Peking University Health Science Center super­visedthe acquisition of data.

Preparationof NGF-loaded microspheres

Figure1. Scanning electronmicrographs of nerve growth factor-loaded mi­crospheres. The microspheres werefabricated by a modified W/O/W emul­sion solvent evaporation method. Poly-D,L-lactide-co-glycolide was the en­capsulation material, and ovalbumin was usedas a protective additive.

NGF-loaded microspheres (Figure 1) were pre­pared by a modified W/O/W emulsion solventevaporation method as published previously. Briefly, a 0.1-ml internal aqueousphase containing 5 μg of 2.5 S NGF (purified from male mouse submaxillaryglands, Promega, USA) and 10 mg of ovalbumin (OVA, Sigma, USA), which was usedas a protective additive, were emulsified in 2 ml of methylene chloridecontaining 50 mg of PLGA (50:50, eta=0.25 dL/g, DURECT, USA). The emulsion wassoni­cated for 30 seconds on an ice bath to create the primary emulsion. Then,under continuous­ly stirring at 1500 rpm, the primary emulsion was addeddropwise into 30 ml of a 3% (w/v) external aqueous solution of polyvinylalcohol (PVA, Sigma, USA) to obtain a multiple emul­sion. After 5 minutes, theresulting emulsion was poured into 300 ml of 0.3% w/v PVA and stirred with amagnetic stirrer for 3 hours at room temperature to evaporate the dichloro­methane.Finally, the resulting suspension was centrifuged, and the collectedmicrospheres were washed with deionized water for three times and freeze-driedto obtain a free-flowing powder. The mean size of the microspheres is 8.1±5.7μm. The protein-loading (w/w) and encapsulation (%) effi­ciency achieved forNGF were 0.0024%, 13.85%, respectively. The initial burst release of NGF frommicrospheres was 18.77%, and 64.34% of the encapsulat­ed NGF was released overa three-week period, as deter­mined by an enzyme-linked immunosorbent assay(ELISA, Promega, USA).

Animals modeland experimen­tal design

A total of 48 male Sprague-Dawley rats, weighting 200-250 g,were used for this study. The animals were housed in trans­parent cages in aSPF facility and given a rodent diet and water ad libitum. Surgical proceduresfor the experimental animals were performed under a binocular surgicalmicroscope using a micro­surgical technique. SD rats were anesthetized with 2%pentobarbitone (0.2-0.3 ml/100 g) by intraperitoneal injection. Afteranesthesia, the right lower limbs were shaved and sterilized. The sciatic nerveand its crotch were exposed and freed from surrounding tissue. Sciatic nerveinjury models were constructed by tran­secting the right sciatic nerve at 7 mmabove the sciatic nerve fork (Figure 2). The animalsreceived a chitin conduit  consisted of polysaccharide shell that demonstrated satis­factory biocompatibility and degradation char­acteristics tobridge a sciatic nerve injury left a 2 mm gap. Conduit size: tube length 4 mm,thickness 0.5 mm, inner diameter 1.5 mm. Both the distal and proximal ends ofthe nerve were inserted approximately 1 mm into the con­duit; the distancebetween the nerve stumps was 2 mm . All operations were performed by the sameinvestigator.

Figure2. Animal model: chitinconduit bridging SD rat sciatic nerve left a 2 mm gap.

The animals were divided into 4 groups ran­domly (12 animals ineach group). The animals were injected into the conduit cavity between thenerve stumps via microinjection with saline (Group A), 20 ng/ml NGF solution (GroupB), saline containing blank PLGA microspheres (Group C), or 40 ng/ml NGF-loadedmicro­spheres (Group D). Each group had two study endpoints, 3 months follow-upand 1 year follow-up.

Sciatic function index

At the two endpoints, the SFI wasused to assess functional recovery. The index print length factor (PLF), toespread factor (TSF), and intermediary to  spread factor (ITF) were recorded forthe normal control foot (NPL, NTS, NIT) and the corresponding experimental foot(EPL, ETS, EIT) for each rat. The SFI was calcu­lated as follows :SFI=-38.3[(EPL-NPL)/NPL]+109.5[(ETS-NTS)/NTS]+13.3[(EIT-NIT)/NIT]-8.8. An SFIof nearly 0 is normal, whereas an SFI of -100 indicates total impairment of thesciatic nerve.

Electrophysiological examination

Electrophysiological assessment wasper­formed at the corresponding endpoints. After anesthesia, the right repairedsciatic nerves were exposed and carefully isolated from the surrounding tissue.

Stimulating bipolar elec­trodes were placed at the proximal and distal repairsite in each group. The recording elec­trode was placed on the nerve shaftdistally, whereas the ground electrode was placed sub­cutaneously. Rectangularpulses (duration 0.1 ms, 0.9 mA) were used to stimulate the repaired nerves.The compound muscle action potential (CMAP) was recorded, and the motor nervecon­duction velocity (MNCV) was obtained simulta­neously by dividing thedistance between the two stimulating sites by the difference in the onsetlatency using an Oxford electrophysiolog­ic instrument.

Weight of triceps surae muscles

After electrophysiologicalassessment, the ani­mals were sacrificed by using an intra-arterial overdose of sodium pentobarbital. Triceps surae muscles from theoperated and non-operated sides were dissected and detached from the bone attheir origin and terminal point, and weighed immediately using an electronicscale. The conserved muscle-mass ratio was recorded for each animal by dividingthe muscle mass on the experimental side by the muscle mass on the control side.

Histologicassessment

The sciatic nerve at2 mm distal to the conduit was removed from each rat, and stained by osmiumtetroxide. The dissected tissues were fixed in 4% paraformaldehyde in 0.1 Mphos­phate buffer for 12 h at 4°C. The fixed nerves were rinsed with runningwater and then rinsed twice in phosphate buffer. The samples distal to theconduit were post-fixed in 1% osmium tetroxide for 12 h. After dehydration andembedding, the specimens were cut into 5-μm cross-sections. The images wereobtained under a Leica microscope (Leica DM 4000B, Germany) at differentmagnifications. Five images from different parts of each section were analyzed,and data from five nerve sec­tions were quantified. Finally, the average thick­nessof the myelin, total number of myelinated axons, and average area of the axonswere evaluated by using a Leica Q550CW analytical system.

Statisticalanalysis

The results wereexpressed as the mean and standard deviation. Differences among groups wereevaluated using t-tests. Analysis was per­formed by using the SPSS version 16.0soft­ware (SPSS Inc., USA). P<0.05 was defined as statistically significant.

Results

General observations

The conduits in all 48 rats werewell-tolerated. The wounds healed without infection, and no trophic ulcerationappeared. Toe spread was observed in all animals, but autophagy was observed inthe right feet of five animals (2 in saline group, 1 in NGF solution group, 1in blank microspheres group and 1 in NGF microspheres group).

Sciaticfunctional index

The SFI of 1 yearsubgroups was significantly improved relative to that of 3 months in all groups(P<0.01). There was no statistically sig­nificant difference among the 3months sub­groups. However, the SFI was significantly improved in NGFmicrospheres group (-26.64±2.24) compared to the saline group (-30.96±3.20,P<0.05) and NGF solution group (-31.97±3.61, P<0.05) at 1 year (Figure 3A).

Figure3. The functionalparameters of the regenerated sciatic nerves at 3 months and 1 year. Thesciatic functional index (A), the motor nerve conduction velocity (B) and theratio of conserved muscle-mass (C). The conserved mus­cle mass ratio wasrecorded for each animal by dividing the muscle mass on the experimental sideby the muscle mass on the control side. The bars represent the mean±SD. n=6,*P<0.05, **P<0.01.

Motor nerveconduction velocity

The results of MNCVat 1 year rats were higher than at 3 months ones in all groups (Figure 3B). The MNCV of NGF microsphere group (31.71±4.55 m/s) was a bithigher than the other groups (29.41±6.14 m/s, P>0.05; 27.90±8.35 m/s,P>0.05; 30.24±5.40 m/s, P>0.05; respectively) at 3 months. There was alsono statistically difference at 1 year. The MNCV of saline group (33.24±3.44m/s) was a bit higher than the other groups at 1 year (32.42±6.24 m/s,P>0.05; 31.76±6.40 m/s, P>0.05; 32.80±8.71 m/s, P>0.05;respec-tively).

Weight oftriceps surae muscles

The ratio ofconserved muscle mass at 1 year was significantly improved compared to that of3 months (P<0.05), except in blank micro­spheres group (P>0.05). Theratio of conserved muscle mass in NGF microsphere group (56.41±4.40%) washigher than the others (54.81±4.15%, P>0.05; 53.05±5.47%, P>0.05;54.47±5.23%, P>0.05; respectively) at 3 months (Figure 3C). The 1 year subgroup of NGF microsphere group (66.40±6.69%) also showed improvementrelative to the other groups (63.31±7.81%, P>0.05; 61.91±5.97%, P>0.05;61.94±8.55%, P>0.05; respectively).

Histomorphometry

The results of the morphometricanalysis of each group are summarized in Table 1 and 2. The myelinated fiber count at 2 mmdistal to the conduit at 1 year was less than at 3 months in all groups(P>0.05). The myelin sheath thick­ness of myelinated nerves of the salinegroup, NGF solution group, blank microspheres group and NGF microspheres groupat 1 year (1.33±0.21 μm, 1.27±0.12 μm, 1.38±0.12 μm, 1.41±0.13 μm,respectively) was obviously improved relative to that at 3 months (0.99±0.12μm, P<0.01; 1.04±0.08 μm, P<0.01; 0.93±0.18 μm, P<0.01; 1.18±0.13 μm,P<0.05; respectively). The axon area of myelinated nerves in all groups at 1year (12.15±0.55 μm2, 12.53±1.43 μm2, 12.47±0.92 μm2, respectively) was obviously improved than at 3 months (10.34±1.13 μm2, P<0.01; 10.29±0.62 μm2, P<0.01; 10.61±0.74 μm2, P<0.01; respectively), except inNGF micro­spheres group (13.02±1.17 μm2 vs. 11.98±1.4 μm2, p>0.05). The fiber count, myelinsheath thickness, axon areaof NGF microsphere group were significantly higher than the saline group(P=0.05, P<0.05, and P<0.05, respectively) at 3 months. However, themorphological param­eters of NGF microsphere group were slightly higher thanthe saline group (P>0.05, P>0.05, and P>0.05, respectively) at 1 year.Either in 3 months or 1 year subgroup, the results of NGF solution and blankmicrosphere groups had no statisticallydifference compared to the saline group (P>0.05) (Figure 4).

Figure 4. Light microscopy images oftransverse sections of the sciaticnerve 2 mm distal to the conduit (stained by osmium tetroxide. x400): Chitin conduit small gap tubulization at 3 months. (A: Saline: B: NGF solution: C: Saline conta ining blank microspheres: D: Saline containing NGF.loaded microspheres) and 1year (E: Saline:F: NGF solu. tion:G: Saline containing blank microspheres: H: Sa line containing NGF.loaded microspheres)

Discussion

Small gap tubulization is based onthe nerve-selective regeneration theory put forward by Cajal . We trieddifferent gap between the two ruptured stumps. After series of experi­ments, wefound that the 2 mm was the most suitable gap in biological degradable conduitin rats 1, 3, 4 and confirmed that the repair effect was advantageous relativeto the traditional epineurial neurorrhaphy. Here, we only used this method torepair nerve injury, but not nerve defect. Based on these results, we weretrying to load exogenous NGF in the con­duit for further improvement ofregeneration effect.

NGF is produced bythe target organs of sen­sory and sympathetic nerves and retrogradely transportedin the neurons. When the peripheral nerve is injured, the injury serves asa signal to stimulate NGF synthesis and promote retrograde transport toneurons. NGF internalization and transport is mediated by high-affinitytyrosin kinase receptors (trkA). The receptors are express on the growthcones surface, and the expression of these receptors in neurons may beincreased by NGF. Sustained release bioactive NGF can increase theexpression of high-affinity tyrosin kinase receptors and accelerate theretrograde transport of NGF to neurons to further support the survival ofembryonic dorsal ganglion neu­rons.

NGF solution is notstable under physiological conditions, so direct administration of NGF into aconduit is difficult to maintain the bioactivity and the effective concentration during nerve regeneration.Polymeric microspheres were able to solve this problem. In several studies , NGF-loaded microspheres were prepared by different encapsu­lationmaterial and combined with the nerve conduits. The results showed that NGFmicro­spheres could improve nerve regeneration to a certain ex-tent. But theobject of those studies was a nerve injury, to explore the use of nerve conduit instead of autologous nerve graft, andthey only obtained short-term results. The effect of small gap tubu­lizationfor peripheral nerve injury combined with NGF sustained release is not clear,espe­cially in long-term effect. For sufficiently long postop observationendpoints for monitoring the progression of the morphological and func­tionalpredictors of recovery, we set two end­points, the three months and 1 yearres-pectively.

The dose of NGF hasalso been studied. Conti et al. demonstrated that maximal neurite outgrowth wasobtained at a concentration of 20 ng/ml. An in vitro study showed an effectiveconcentration of NGF and found that overdos­ing would inhibit outgrowth. So theconcen­tration of NGF solution was set at 20 ng/ml. In vitro experimentsconfirmed that only approxi­mately half of the encapsulated NGF was releasedfrom the NGF-loaded microspheres. Therefore, the NGF microspheres were set toprovide a NGF concentration of 40 ng/ml.

In this study, thebase material of the micro­spheres was PLGA, which is a biodegradable materialthat has been widely used in polymeric microspheres, and the ratio of lactic toglycolic acid can be modulated to alter the release characteristics. In thestudy, the groups either with 3 months or 1 year follow-up, the result of theblank microsphere group was not worse than the saline group, demonstrating thatthe PLGA microsphere delivery system did not inhibit or obstruct the outgrowthof nerve fibers.

Myelin sheaththickness and myelinated nerve axon area provides a measure of the maturity ofthe regenerating nerve fibers. At 3 months, the maturity of the myelinated nerves with NGF sustained releasewas much better than the absence of exogenous NGF or direct adminis­tration ofexogenous NGF. This result indicated that sustained release NGF was importantto promote maturation of fibers in the distal nerve. This phenomenon may be dueto that the released NGF binds to the p75 low-affinity NGF receptor on Schwanncells and up-regulates its expression, which promotes the migra­tion of Schwanncells into the gap between the injured stumps32. This helps the formation ofBüngner bands in basal lamina tubes and stim­ulates myelinization.

Maturity ofmyelinated nerve fibers at 1 year was superior to 3 months. The results were inagreement with those of Fox IK et al. Obviously, with the extension of the timeafter operation, Myelin sheath thickness and axon area would be improved. Inaddition, those wrong matching nerve fibers which could not reach the targetorgan, often relatively imma­ture and eventually pruned. Consequently, theaverage value of the param­eters of the effective nerve fibers retained evenbetter. The myelin sheath thickness and axon area did not show significantdifferences among the four groups at 1 year, in contrast to out­standingperformance of the NGF microspheres group at 3 months. In other words, even ifthere was no sustained release of NGF, nerve fibers could reach a certainmaturity, but require more time.

It is interested tonote that the total amount of myelinated nerve fibers number at 1 year post­opwas lower than the fibers number at 3 months.The results of this study were in agreement with those of Muratori L. Thisobservation could be inter­preted as the result of multiple axonal sproutsduring the early regeneration phases followed by a late pruning of the branchesthat were wrong matching and could not reach their target organ.

Because earlymyelinated nerve fibers are more mature, the SFI of NGF sus­tained releasegroup was better than the other groups. With the extension of the time afteroperation, the better maturity of the nerve fibers establish­ing functionalconnections distally made SFI apparentrecovery, especially in NGF sustained release group which indicated thatsustained release NGF improved functional recovery in the long term. Thisphenomenon may be the result of the early NGF sustained release and play therole on regeneration. But earlier or later, the NGF microspheres had little orno influence on MNCV or the ratio of con­served muscle mass. It was notdifficult to associate to previous studies found that NGF has little or noinfluence on motoneurons and their neurite outgrowth, that was thelimitations of this study. Future studies could focus on microspheresassociated with other various neurotrophic factors for nerve regener­ation andthe evaluation of relevant neurons and target organs.

In conclusion, the release of NGF microspheres in small gaptubulization benefit on peripheral nerve injury facilitated nerve regenerationhis­tologically, especially for the maturity of early regenerative myelinatednerve fibers and also had an effect on functional recovery in the long term.

参考文献（略）