lable at ScienceDirect

Biomaterials 31 (2010) 1596–1603

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

Polyester based nerve guidance conduit design

Deniz Yucel a, Gamze Torun Kose b, Vasif Hasirci a,c,d,e,*

a METU, BIOMAT, Department of Biotechnology, Biotechnology Research Unit, Ankara 06531, Turkeyb Yeditepe University, Department of Genetics and Bioengineering, Faculty of Engineering and Architecture, Istanbul 34755, Turkeyc METU, BIOMAT, Department of Biological Sciences, Biotechnology Research Unit, Ankara 06531, Turkeyd METU, BIOMAT, Department of Biomedical Engineering, Biotechnology Research Unit, Ankara 06531, Turkeye METU, BIOMAT, Department of Micro and Nanotechnology, Biotechnology Research Unit, Ankara 06531, Turkey

a r t i c l e i n f o

Article history:Received 1 September 2009Accepted 3 November 2009Available online 22 November 2009

Keywords:Nerve regenerationNerve guideConduitElectrospun matMicropattern

* Corresponding author. Middle East Technical Uniof Biological Sciences, Biotechnology Research Unit,Turkey. Tel.: þ90 312 210 5180; fax: þ90 312 210 154

E-mail addresses: dyucel@metu.edu.tr (D. Yucel)(G.T. Kose), vhasirci@metu.edu.tr (V. Hasirci).

0142-9612/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.biomaterials.2009.11.013

a b s t r a c t

Nerve conduits containing highly aligned architecture that mimics native tissues are essential for effi-cient regeneration of nerve injuries. In this study, a biodegradable nerve conduit was constructed byconverting a porous micropatterned film (PHBV–P(L-D,L)LA–PLGA) into a tube wrapping aligned elec-trospun fibers (PHBV–PLGA). The polymers were chosen so that the protective tube would erode slowerthan the fibrous core to achieve complete healing before the tube eroded. The pattern dimensions andthe porosity (58.95 (%) with a maximum pore size of 4–5 mm) demonstrated that the micropatterned filmwould enable the migration, alignment and survival of native cells for proper regeneration. This film hadsufficiently high mechanical properties (ultimate tensile strength: 3.13 MPa, Young’s Modulus: 0.08 MPa)to serve as a nerve guide. Electrospun fibers, the inner part of the tubular construct, were well alignedwith a fiber diameter of ca. 1.5 mm. Fiber properties were especially influenced by polymer concentration.SEM showed that the fibers were aligned parallel to the groove axis of the micropatterned film within thetube as planned considering the nerve tissue architecture. This two component nerve conduit appears tohave the right organization for testing in vitro and in vivo nerve tissue engineering studies.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Nerve damage could result from mechanical, thermal, chemical,or ischemic factors and could lead to the disruption of thecommunications between neurons and their supporting cells [1].The annual incidence of spinal cord injury in the USA is approxi-mately 12,000 [2] and five percent of all open wounds in theextremities caused by accidents are complicated by peripheralnerve trauma [3]. Self regeneration of nervous tissue is difficult insevere damages of peripheral nervous system (PNS) and almostimpossible in central nervous system (CNS) [4,5]. In PNS injuriesend-to-end anastomosis is a commonly used technique for smallgaps to bridge the severed nerve ends via suturing [6]; however, itbecomes difficult for long nerve gaps without applying any tension[7]. In such cases the most widely used technique is the use ofautologous nerve grafts, grafting of a nerve segment removed from

versity, BIOMAT, DepartmentInonu Bulvari, Ankara 06531,2.

, gamzekose@yeditepe.edu.tr

All rights reserved.

another part of the patient’s body. Despite successful results theseautologous grafts have inherent disadvantages, such as limitedsupply, permanent loss of the nerve function at the donor site andneed for multiple surgeries [8,9].

Development of alternative treatments, especially for largerdefects, is necessary to bridge the gap between the proximal andthe distal nerve stumps. Biological or synthetic tubular nerveconstructs with highly aligned architecture mimicking the nativetissue could provide the bridge needed for nerve regeneration[10–13]. Directional axonal elongation is mainly based on theinteractions between regenerating axons and the adjacentsubstratum [14]. Incorporation of some special micro and nano-architectures that allow structural support for axonal regrowth andaffect cellular orientation is a promising strategy in nerve conduitdesign [15]. The performance of the nerve guides may be improvedby the use of desired surface texture, longitudinally-orientedmicrochannels or polymer fibers [16]. The most common techniqueto produce patterned surfaces with controlled dimensions andspecific shapes is microfabrication [17], by the use of photolithog-raphy which can be followed by micromachining, etching ordeposition [18]. When the native neuron orientation and migrationalong the anisotropic direction is taken into account patternedsubstrates with channels appear to be more suitable for the control

D. Yucel et al. / Biomaterials 31 (2010) 1596–1603 1597

of neural cell orientation [19–22]. Fibrillar scaffolds, on the otherhand, are generated by pressure-assisted microsyringe, selfassembly, and electrospinning techniques. Use of uniaxiallyoriented, biodegradable electrospun fibrous mats is a promisingapproach to the restoration of the damaged nerve because itmimics the native architecture of the nerve tissue, and directionalcell growth is a prerequisite for functional nerve regeneration.

Ideally, the nerve conduit should be porous to allow andcontrol nutrient exchange and biodegradable to eliminate theneed for its removal [23]. In this regard the choice of theconstruct material becomes an important point. Extensively usedsynthetic polymers, including polylactic acid (PLA) [24] andpoly(D,L-lactide-co-glycolide) (PLGA) [25] are known for theirease of processing, low inflammatory response, and approval bythe U.S. Food and Drug Administration. Naturally derived poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) is a degradableand biocompatible polymer of natural origin used in the design ofimplants. Its degradation products, 3-hydroxybutyric acid, whichis a normal constituent of blood [26], and 3-hydroxyvaleric acid,are not known to lead to any long term tissue reaction at theimplantation site.

In the present study, a polyester based, biodegradable, porousnerve guidance conduit was constructed. The conduit was designedfor eventual use in the regeneration of nerves across long nervegaps. The construct was composed of two parts; an aligned, elec-trospun mat of PHBV and PLGA serving as the inner component anda porous, micropatterned film of PHBV, poly(L-lactide-co-D,

Fig. 1. Scanning electron micrographs of micropatterned polymeric films obtained from thew/w) film, (b) porous micropatterned PHBV–P(L-D,L)LA–PLGA (2:2:1, w/w) film. (c–d) Cross-the film (left) is micropatterned (MP) bottom, and the other side (right) is smooth top. Insetsis the image of the surface, and inset 2 is the image of the bulk just beneath the surface (c

L-lactide) (P(L-D,L)LA) and PLGA as the tubular, outer part designedto wrap around the electrospun mat. The surface topography andmacroscopic characteristics of the porous micropatterned films andthe aligned electrospun mats were examined with scanning elec-tron microscopy (SEM). The mechanical properties of the filmshowed that it is not rigid and could be sutured. The erosion rate ofthe fibrous mat and the micropatterned film was studied. In thisstudy, the biodegradability of the polymers, the porosity of the film,and the channels for physical guidance satisfy the requirements foran ideal nerve conduit. In addition, the use of two different topo-graphical cues, electrospun fibers and micropatterned films, ina single design is expected to maximize the influence of the guid-ance cues. Thus, the construct prepared in this study could beconsidered for use as a conduit in nerve regeneration and in therepair of gaps longer than a few centimeters by promoting thealignment of neurons and supporting cells.

2. Materials and methods

2.1. Preparation of polymeric films

Micropatterned (MP) silicon (Si) templates with perpendicular walls wereproduced by photolithography and subsequent reactive ion etching (kindly providedby Prof. A. Aydınlı and A. Kocabas, Bilkent University). A negative polydimethylsiloxane (PDMS) replica, obtained by using PDMS prepolymer–catalyst mixture(Sylgard 184 Elastomer Kit, Dow Corning, U.S.A), served as the template to preparethe micropatterned polymeric film via solvent casting. A PHBV (5% by mole of 3-hydroxyvalerate, Fw: 222.2 g/mol, Aldrich, UK), P(L-D,L)LA (70:30, InherentViscosity: 5.5–6.5 dL/g, AppliChem, Germany) and PLGA (50:50, Inherent Viscosity:

PDMS replica of the Si template. (a) Nonporous micropatterned PHBV–P(L-D,L)LA (1:1,sections of porous micropatterned films ((d) at higher magnification). In (c) one side ofof (c) confocal micrographs of porous micropatterned film stained with Nile Red, inset 1a. 10 mm deep).

Table 1Tensile properties of various polymeric films.

Sample UTS (MPa) E (MPa)

PHBV5–P(L-D,L)LA 29.16� 1.23 1.54� 0.01PHBV5–P(L-D,L)LA–PLGA 20.81� 1.47 1.05� 0.23Porous PHBV5–P(L-D,L)LA–PLGA 3.13� 0.44 0.08� 0.02

UTS: ultimate tensile strength, E: modulus of elasticity or Young’s Modulus.

Fig. 2. Scanning electron micrographs of PHBV–PLGA fibers prepared under constantpotential (18 kV) and flow rate (20 mL/min) (a) with a 10% (w/v) polymer solution anda distance of 15 cm and (b) with a 15% (w/v) polymer solution and a distance of 25 cm.Insets are images taken at higher magnifications.

D. Yucel et al. / Biomaterials 31 (2010) 1596–16031598

0.32–0.44 dL/g, Boehringer Ingelheim Pharma KG, Germany) (2:2:1, w/w) solution inchloroform (5%, w/v) was prepared. Polyethylene glycol (PEG) (10%, w/v) was addedto this solution to serve as a porogen because of its low molecular weight (1300–1600 Da) compared to the other used polymers. Polymer solution was poured on thePDMS template, and following solvent evaporation the film was peeled off thesurface. The films were then repeatedly washed in distilled water to extract PEGsince it was soluble in water while the others were not, and thus a porous micro-patterned film was obtained.

2.2. Preparation of electrospun fibrous mats

Microfibers were obtained by electrospinning using a collector made of twoparallel metal rods spaced 40 mm apart, and a solution of PHBV–PLGA (1:1 w/w, inchloroform:N,N-dimethylformamide (DMF)) as the polymer solution. Certainparameters such as concentration of polymer solution (5%, 10% and 15%, w/v), thedistance between the two poles (15–25 cm), the flow rate of the polymer solution(10–30 mL/min) and the potential (15–25 kV) were varied during the optimizationprocess.

In order to study the uniformity of PLGA and PHBV and their distribution in thefiber the mat was immersed in acetone for 2 h to dissolve out PLGA thus leavingbehind PHBV.

2.3. Preparation of 3D nerve guidance conduit

The porous, micropatterned film (30 mm in length and 10 mm in width) wasrolled around the electrospun mat to form a 3D tubular structure and was main-tained in this form using an acrylate-based adhesive.

2.4. Characterization

2.4.1. Scanning electron microscopySurfaces of the samples were coated with Ag–Pd under vacuum and SEM was

carried out by a QUANTA 400F Field Emission SEM at 10 kV.

2.4.2. Porosity of the polymeric filmsThe porous PHBV–P(L-D,L)LA–PLGA films obtained after leaching out the PEG

were stained with a hydrophobic dye, Nile Red, and then were examined usingconfocal microscopy (Leica-DM 2500) and the images of three different regions ontwo separate micropatterned films were taken in 1 mm layers in z-direction. Theporosity of the micropatterned films was assessed using confocal microscopy andScion Image (NIH) program.

2.4.3. Mechanical analysis of the polymeric filmsMechanical properties of polymeric films were studied in a wet state at room

temperature by Lloyd LRX 5K Mechanical Tester, controlled by the program WindapR.Films (40.0 mm long, 10.0 mm wide) of PHBV–P(L-D,L)LA, PHBV–P(L-D,L)LA–PLGA,and porous PHBV–P(L-D,L)LA–PLGA (thickness 0.028� 0.006 mm, 0.072� 0.007 mm,and 0.113� 0.012 mm, respectively) were attached to the clamps (gauge length:10 mm) of the instrument and tested at a rate of 10 mm/min.

2.4.4. Solid state nuclear magnetic resonance spectroscopy (NMR)The removal of PLGA upon acetone treatment was studied by the solid state

NMR spectroscopy. The chemical characterization of the electrospun fibrous matsbefore and after acetone treatment was done by High Power Solid State 300 MHzNMR Spectrometer (Bruker, Superconducting FT.NMR Spectrometer Avance�, with300 MHz Wide Board Magnet) running 13C CPMAS analysis at a spin rate of 5000 Hz.The powder of PHBV and PLGA was used as a reference to obtain the characteristicpeaks of each polymer separately.

2.4.5. Erosion of micropatterned films and electrospun matsPHBV–P(L-D,L)LA, PHBV–P(L-D,L)LA–PLGA films (porous and nonporous) and

PHBV–PLGA mats (10� 30 mm2) were dried completely and weighed (ca. 80 mg forfilms and 50 mg for mats). The samples were incubated in phosphate buffered saline(PBS, 30 mL, pH 7.4, 10 mM) at 37 �C for 90 days with continuous shaking. At varioustime points samples were removed from PBS, dried completely under vacuum atroom temperature, weighed and then placed back in the same PBS solution tocontinue the study.

3. Results and discussion

3.1. Surface topography of micropatterned polymeric films

The micropatterned and perpendicular walled Si templateswere produced via photolithography followed by reactive ionetching. The patterns on its PDMS replica were successfullytransferred onto the biodegradable polymeric films by solventcasting (Fig. 1a and b). The dimensions of the polymeric films(groove width (GW): 14–14.5 mm, ridge width (RW): 5 mm, groovedepth (GD): 5 mm) were practically the same as those of the PDMStemplate (GW: 5 mm, RW: 15 mm, GD: 5 mm). The ridges of thePDMS template became the grooves on the film.

The porous, micropatterned polymeric film was designed toserve as the exterior of the tubular nerve construct. Thus, it had to

Fig. 3. SEM of PHBV (5%, w/v) fibers. (a) untreated, and (b) after 2 h acetone treatment. SEM of PHBV–PLGA (15%, w/v) fibers prepared under optimized conditions: (c) beforeacetone treatment, (d) after a short acetone wash, (e) a higher magnification of (d) (droplets and streaks are indicated by circles and arrows, respectively), (f) after 2 h of acetonetreatment (the circles show the nicks). 13C CPMAS NMR spectrum of PHBV–PLGA fibrous mat (g) before and (h) after acetone treatment.

D. Yucel et al. / Biomaterials 31 (2010) 1596–1603 1599

be permeable to the nutrients and waste products to allow cellgrowth within the construct. It also had to be resilient in order tobe able to secure the 3D tubular construct to the native nervetissue, therefore, PLGA was added to the composition to improvefilm flexibility. The other important criterion, porosity, could beintroduced and/or increased by addition of some low molecularweight additives, that could later be leached out [25]. In this study,this was achieved by addition of PEG to the film composition andthen leaching it in distilled water. It can be seen from the SEMimages that the films obtained after PEG removal are fairly porousand maintained undisrupted micropatterns (Fig. 1b). The porousstructure on the surface can also be seen in the bulk of the filmeven though the skin layer appears to be less porous than the rest(Fig. 1c and d).

3.2. Porosity of the polymeric films

The porosity of the PHBV–P(L-D,L)LA–PLGA films was deter-mined by using the confocal microscopy images of the micro-patterned films followed by analysis with the NIH Scion Imageprogram. Total porosity (including the surface and the bulk of thefilm) was calculated by using the images of layers (1 mm inz-direction) of three different regions of two separate micro-patterned films (insets given in Fig. 1c). The porosity was deter-mined as 58.95� 2.42 (%). In another study of PLGA nerve guidancechannels, a similar level of porosity (�50%) was obtained by the useof P103 porogen [25].

In the SEM image in Fig. 1b the maximum pore size observed onthe patterned film is around 4–5 mm, and therefore, it is smaller

D. Yucel et al. / Biomaterials 31 (2010) 1596–16031600

than the size of most neural cells. Similarly, a 1–10 mm pore size wasreported in the construction of nerve guides [27]. Thus, this level ofpore size on the micropatterned film would allow the nutrients topermeate, however, would not allow the permeation of the cellsforcing the cells to remain in the tube, align and migrate along theaxis of the micropatterns. This is in compliance with the approachin this study because the role of this film was to align the cells andmaintain an environment suitable for growth within the construct.

Fig. 4. Erosion profile of (a) electrospun mat (PHBV–PLGA) and (b) polymeric films.SEM images of PHBV–PLGA fibers before and after erosion are given as insets in theerosion plot of the electrospun mat.

3.3. Tensile properties of the polymeric films

The tubular construct was designed considering the eventualuse, in that it would be connected to the distal and proximal ends ofthe native nerve tissue through suturing. Thus the outer (micro-patterned film) region had to be able to withstand the tensile stresscreated during suturing.

In the tensile tests, the modulus of elasticity and ultimate tensilestrength of three different polymeric samples were determined(Table 1). The results showed that PLGA incorporation into PHBV–P(L-D,L)LA film led to ca. 30% decrease in both UTS and elasticmodulus. These indicate that PLGA incorporation made the filmweaker, softer but more extensible. Upon incorporation of porosityUTS and elastic modulus were decreased substantially (85% and92%, respectively). These indicate that the mechanical properties ofthe graft could be fine tuned by incorporating porosity, and tolesser extent by introduction of PLGA, to the blend composition.Similarly, it was reported that the tensile mechanical properties ofa conduit were changed upon incorporation of uniform porosity; itled to a decrease in maximum stress due to loss of their compactinternal microstructure [28].

In the literature the UTS and elastic modulus for normal nerveand for acellular nerve are given as 2.7 MPa and 0.580 MPa, and as1.4 MPa and 0.576 MPa, respectively [29]. Since the conduit con-structed in this study is initially free of cells it is more appropriateto compare it with the acellularized nerve. According to the valuesgiven in Table 1 the outer (porous film) part of the present conduitis stronger than both the acellularized and the normal nerve.Moreover, the lower elastic modulus of the construct shows that itis also less rigid than both acellularized and normal nerves. It canthus be stated that the construct at hand has sufficiently highmechanical properties to serve as a nerve guide.

3.4. Physical and chemical characterization of the electrospunfibrous mat

3.4.1. Fiber formationIn the preparation of electrospun fibers various process

parameters were modified to obtain uniform, defect-free, andaligned fibers. To improve the fiber quality DMF (has a highdielectric constant, 36.7 at 25 �C), was added to the polymer solu-tion. According to the literature, by increasing the conductivity ofthe polymer solution, DMF would help form more straight andmore uniform fibers without any fusion or beads [30].

Among the parameters varied, polymer concentration wasfound to be very influential on fiber diameter. It was not possible toobtain a PHBV–PLGA fibrous mat using a low polymer concentra-tion, 5% (w/v), even when the DMF content was high. It waspossible to obtain aligned PHBV–PLGA fibrous mats with fibers950 nm in diameter (with some beads) with a 10% (w/v) solution(Fig. 2a). The best mats were obtained when the concentration was15% (w/v) (Fig. 2b). Electrospun fibers produced under optimizedconditions of (15% w/v) polymer solution, 18 kV potential applied,20 mL/min flow rate, 25 cm the distance were well aligned, withoutbeads and with a fiber diameter of ca. 1.5 mm (Fig. 2b).

3.4.2. Fiber component distribution study with SEM and NMRThe solvent used in dissolving the polymer blends was chloro-

form and it was not an equally good solvent for all the components.It would, therefore, be possible that the distribution of thecomponents in a fiber may well be anisotropic due to phase sepa-ration. In the present case to study the possibility of phase sepa-ration acetone was used to dissolve out one of the components.PLGA is soluble in acetone whereas PHBV is not; therefore, thefibers were treated with acetone to study the homogeneity of PLGAand PHBV distribution within the mat (Fig. 3). The acetone treat-ment was not effective on PHBV fibers (Fig. 3a and b). However,droplet-like semispherical formations were observed on the fibersof this blend (PHBV–PLGA) after a short acetone treatment prob-ably due to initiation of PLGA dissolution (Fig. 3e). After 2 h ofacetone treatment PLGA dissolved out leaving behind modifiedsurfaces and nicks (Fig. 3f). These SEM results showed that PHBV–PLGA fibers do not have core–shell structure. The streaks, seen inFig. 3e, indicate that PLGA is generally found mixed with PHBVthroughout the fiber; however, the frequent breaks (nicks) indicateincomplete mixing of the two polymers in some regions of thefibrous mats (probably PLGA-rich regions broke). The change in thefiber structure was also seen in the diameter of fibers which wasdecreased from ca. 1.5 mm to ca. 1 mm upon acetone treatment.

The 13C CPMAS NMR spectrum of the fibers before and afteracetone treatment supported the SEM images in showing completeremoval of PLGA from the PHBV–PLGA fibrous mat upon acetonetreatment (Fig. 3). The peaks, which are characteristic for PLGA, at52 ppm and 8 ppm (seen as a shoulder) correspond to CH2 in gly-colic acid and methylene groups of D,L-lactic acid, respectively,

Fig. 5. Formation of 3D construct and (a–b) various cross sectional views of the construct by SEM (EM: electrospun mat, MPF: porous micropatterned film). Inset 1 of a: SEM imageof the exterior surface of the tubular construct. Inset 2 of a and an inset of b: SEM images of certain parts of the constructs at higher magnifications.

D. Yucel et al. / Biomaterials 31 (2010) 1596–1603 1601

disappeared after acetone treatment by the removal of PLGA(Fig. 3g and h). However, the peaks, observed in PHBV (data notgiven), at 59 ppm, 33 ppm and 10 ppm were present both beforeand after acetone treatment. Thus, NMR results confirm that thestreaks seen in SEM images of acetone treated fibers were due toremoval of the PLGA component from the mats, and the remainderwas basically PHBV.

3.5. Erosion of polymeric films and electrospun mats

The disappearance of an implant with time is an importantproperty to be considered in the design of a biomaterial to elimi-nate the risk of long term presence of degradation products andalso the need for second surgery to remove the implant. Thus, theerosion rates of electrospun fibrous mats of PHBV–PLGA and filmsof PHBV–P(L-D,L)LA, PHBV–P(L-D,L)LA–PLGA and porous PHBV–P(L-D,L)LA–PLGA were investigated gravimetrically (Fig. 4).

The weight of PHBV–PLGA electrospun fiber mats decreasedsteadily with time (Fig. 4a). At the end of 90 days there was an 80%decrease in the sample weight which was probably due to thecomplete loss of PLGA and loss of a fraction of PHBV as expected fromthe degradation rate of PLGA being higher than that of PHBV [31,32].The PLGA loss would be due to degradation and erosion, while PHBVloss would probably be due to erosion. SEM images given as insets inFig. 4a also support this statement. Upon degradation of electrospunmats the streaks were formed on the fibers like those seen in Fig. 3f asPLGA removal after acetone treatment, and the dimension of the fiberwas decreased in diameter from ca. 1.5 mm to ca. 1.1 mm.

The weight of PHBV–P(L-D,L)LA films remained almost the samefor the whole duration (5% decrease in weight in 90 days) (Fig. 4b).However, incorporation of PLGA increased the erosion rate of thefilms. A sharp (15%) decrease in weight was observed between days20 and 30, and the total weight loss in 90 days was around 35%.Surprisingly the porous PHBV–P(L-D,L)LA–PLGA films had slowererosion rates (with a total weight loss of ca. 25%) compared to

nonporous ones. A faster erosion in porous structures was expecteddue to higher penetration of the solvent (water). However, if bulkerosion with PLGA takes place via autocatalysis (due to the degra-dation product, lactic acid and larger molecules, accumulating andspeeding up degradation) then it is possible that the porous samplemay avoid autocatalysis. On the whole, the films maintained theirintegrity for a period of 3 months indicating that during this periodthe core of the 3D construct would be eroded gradually to allownerve regeneration while the protective tube would still be aroundas expected.

3.6. Formation of 3D nerve guidance conduit

The two components of the 3D tubular construct were theexternal porous, micropatterned film and the inner aligned elec-trospun fibrous mat. The conduit was prepared by placing the maton the patterned film, and then rolling with the micropatternsfacing the inside (Fig. 5). The tubular form was maintained by theaid of an acrylate adhesive. The localization of the electrospun matand the porous micropatterned film in the construct can be seen inthe cross sectional views of the conduit (Fig. 5a and b). The inside ofthe porous film which formed the tubular structure can be seenclearly in the inset of Fig. 5b and was micropatterned, while theexterior surface of the construct was smooth (inset 1 of Fig. 5a).SEM images show that the aligned fibers were parallel to the axis ofthe micropatterned film (the axis direction is indicated by the twosided arrows in the inset of Fig. 5b).

Topographical cues such as multi channels [13,33], micron-scale grooves [11,34] and fibers [10,12,35] have been commonlyused as internal structures to mimic the architecture of the nervefascicle for achieving alignment of supporting cells and neuronsand to promote nerve regeneration. In most of the studiesreported a single topographical feature was used. In this study,however, the use of two oriented components, an aligned elec-trospun fibrous mat (inner part) and a porous micropatterned

D. Yucel et al. / Biomaterials 31 (2010) 1596–16031602

film (outer, tubular part), in the same structure was expected tomaximize the topographic directional cues for the migration andorganization of nerve cells along the axis of the fibers and groovesof the films. This alignment of cells is very critical in the regen-eration of the injured nerve tissue. Another advantage of thedesign was that placing the thin and highly porous electrospunmat inside the tubular film would make a large space available forthe cells to grow in.

The length of the final construct was 30 mm, and this could beincreased if needed. The diameter of the final construct was ca.2.5 mm and this could also be changed depending on the size of theinjured nerve. The thickness of the outer layer, the porous micro-patterned film, was determined as 80–100 mm. In several studies thedesigned nerve construct had a similar thickness of 50–100 mm [7,36].The ability to customize the properties of the construct eliminates thedimension mismatch, and would minimize possible adaptationproblems between the injured nerve and the nerve conduit [10].

The approach used in this study is similar to the constructsdescribed recently in the literature and it was revealed that theresults of the in vivo studies of these constructs in promoting nerveregeneration were promising [10,37]. In one study, the non-degradable poly(acrylonitrile-co-methylacrylate) (PAN-MA) elec-trospun fibrous mats were packed in a polysulfone nerve conduit tobuild a 3D polymeric construct [10]. Anatomical and functionalevaluations showed that only the polymeric, aligned fibrousconstructs (but not the random fibers) significantly enhanceperipheral nerve regeneration in 17 mm nerve gaps in rats. Thepresence of non-degradable components in the conduit had noadverse effects on nerve regeneration, but it was indicated that thenext generation of biodegradable nerve guidance conduits, like inour design, would enable regeneration, provided that their degra-dation does not interfere with the stability of the structures formedduring the initial regeneration period. In the other study, a collagenbased nerve conduit composed of a tubular scaffold filled withlongitudinal filaments was tested in the regeneration of 30 mmnerve defects [37]. Even though collagen promoted cellular prolif-eration and tissue healing, its degradation was faster than that ofthe other biodegradable materials such as PGA or PLLA and thiswould be a problem in the case of a long gap (>30 mm).

Consequently, the construct described in this study is expectedto have a good potential to serve as a nerve conduit for regener-ation and functional recovery in nerve tissue engineering appli-cations due to its biodegradability, porosity as well as its highlyaligned, organized micro-architecture of each component of theconstruct.

4. Conclusion

A 3D, biodegradable, porous, polymeric nerve guidance conduitwas described for use in the restoration of the function of injurednerve tissues. In this design two components of the construct, theporous, micropatterned film as the exterior part and the alignedelectrospun fibrous mat as the interior part, are expected tomaximize the effect of topographic cues for the migration andalignment of supporting cells and neurons to enhance the healingprocess. These physical cues are expected to influence endogenousnerve repair mechanisms in the absence of exogenous neurotrophicor extracellular matrix proteins. The addition of chemical cues suchas coating with an extracellular matrix protein would furtherimprove the effectiveness of this conduit.

Acknowledgements

The authors gratefully acknowledge the support of State Plan-ning Organization of Turkey (DPT) through the Project BAP

01.08.DPT.2003K120920-20, METU through the Project BAP-2007-07-02-00-01, FP6 European Network of Excellence project EXPER-TISSUES, The Scientific and Technical Research Council of Turkey(TUBITAK) Nanobiomat 105T508, TUBITAK SBAG 2723, and TUBI-TAK BIDEB (BDP).

Appendix

Figures with essential color discrimination. Figs. 1–3 in thisarticle have parts that are difficult to interpret in black and white.The full colour images can be found in the on-line version, at doi:10.1016/j.biomaterials.2009.11.013.

References

[1] Zhang N, Yan H, Wen X. Tissue engineering approaches for axonal guidance.Brain Res Rev 2005;49(1):48–64.

[2] National Spinal Cord Injury Statistical Center. Spinal cord injury facts andfigures at a glance 2009, Birmingham, Alabama, U.S.A. Online. Available fromURL: http://www.spinalcord.uab.edu; 2009.

[3] Huang YC, Huang YY. Biomaterials and strategies for nerve regeneration. ArtifOrgans 2006;30(7):514–22.

[4] Meek MF, Varejao AS, Geuna S. Muscle grafts and alternatives for nerve repair.J Oral Maxillofac Surg 2002;60(9):1095–6.

[5] Schmidt CE, Leach JB. Neural tissue engineering: strategies for repair andregeneration. Annu Rev Biomed Eng 2003;5:293–347.

[6] Sanghvi AB, Murray JL, Schmidt CE. Tissue engineering of peripheral nerve. In:Wnek GE, Bowlin GL, editors. Encyclopedia of biomaterials and biomedicalengineering. New York: Marcel Dekker Inc; 2004. p. 1613–21.

[7] Ichihara S, Inada Y, Nakada A, Endo K, Azuma T, Nakai R, et al. Development ofnew nerve guide tube for repair of long nerve defects. Tissue Eng Part CMethods 2009;15(3):387–402.

[8] Millesi H. Indications and techniques of nerve grafting. In: Gelbertman RH,editor. Operative nerve repair and reconstruction. Philadelphia: JB LippincottCompany; 1991. p. 525–44.

[9] Terzis JK, Sun DD, Thanos PK. History and basic science review: past, present,and future of nerve repair. J Reconstr Microsurg 1997;13(3):215–25.

[10] Kim Y, Haftel VK, Kumar S, Bellamkonda RV. The role of aligned polymer fiber-based constructs in the bridging of long peripheral nerve gaps. Biomaterials2008;29(21):3117–27.

[11] Lin YL, Jen JC, Hsu SH, Chiu IM. Sciatic nerve repair by microgrooved nerveconduits made of chitosan–gold nanocomposites. Surg Neurol 2008;70(Suppl.1):9–18.

[12] Wang HB, Mullins ME, Cregg JM, Hurtado A, Oudega M, Trombley MT,et al. Creation of highly aligned electrospun poly-L-lactic acid fibers fornerve regeneration applications. J Neural Eng 2009;6(1):016001 [Epub2008 Dec 22].

[13] He L, Zhang Y, Zeng C, Ngiam M, Liao S, Quan D, et al. Manufacture of PLGAmultiple-channel conduits with precise hierarchical pore architectures and invitro/vivo evaluation for spinal cord injury. Tissue Eng Part C Methods2009;15(2):243–55.

[14] Letourneau PC. Cell-to-substratum adhesion and guidance of axonal elonga-tion. Dev Biol 1975;44(1):92–101.

[15] Recknor JB, Sakaguchi DS, Mallapragada SK. Directed growth and selectivedifferentiation of neural progenitor cells on micropatterned polymersubstrates. Biomaterials 2006;27(22):4098–108.

[16] Schlosshauer B, Lietz M. Nerve guides. In: Wnek GE, Bowlin GL, editors.Encyclopedia of biomaterials and biomedical engineering. New York: MarcelDekker Inc; 2004. p. 1043–55.

[17] Flemming RG, Murphy CJ, Abrams GA, Goodman SL, Nealey PF. Effects ofsynthetic micro- and nano-structured surfaces on cell behavior. Biomaterials1999;20(6):573–88.

[18] Li N, Tourovskaia A, Folch A. Biology on a chip: microfabrication forstudying the behavior of cultured cells. Crit Rev Biomed Eng 2003;31(5–6):423–88.

[19] Rajnicek AM, Britland S, McCaig CD. Contact guidance of CNS neurites ongrooved quartz: influence of groove dimensions, neuronal age and cell type.J Cell Sci 1997;110:2905–13.

[20] Mahoney MJ, Chen RR, Tan J, Saltzman WM. The influence of microchannelson neurite growth and architecture. Biomaterials 2005;26(7):771–8.

[21] Goldner JS, Bruder JM, Li G, Gazzola D, Hoffman-Kim D. Neurite bridgingacross micropatterned grooves. Biomaterials 2006;27(3):460–72.

[22] Song M, Uhrich KE. Optimal micropattern dimensions enhance neuriteoutgrowth rates, lengths, and orientations. Ann Biomed Eng2007;35(10):1812–20.

[23] Griffith LG. Emerging design principles in biomaterials and scaffolds for tissueengineering. Ann N Y Acad Sci 2002;961:83–95.

[24] Evans GR, Brandt K, Widmer MS, Lu L, Meszlenyi RK, Gupta PK, et al. In vivoevaluation of poly(L-lactic acid) porous conduits for peripheral nerve regen-eration. Biomaterials 1999;20(12):1109–15.

D. Yucel et al. / Biomaterials 31 (2010) 1596–1603 1603

[25] Wen X, Tresco PA. Fabrication and characterization of permeable degradablepoly(DL-lactide-co-glycolide) (PLGA) hollow fiber phase inversion membranesfor use as nerve tract guidance channels. Biomaterials 2006;27(20):3800–9.

[26] Holland SJ, Jolly AM, Yasin M, Tighe BJ. Polymers for biodegradable medicaldevices. II. Hydroxybutyrate–hydroxyvalerate copolymers: hydrolytic degra-dation studies. Biomaterials 1987;8(4):289–95.

[27] Vleggeert-Lankamp CL, de Ruiter GC, Wolfs JF, Pego AP, van den Berg RJ,Feirabend HK, et al. Pores in synthetic nerve conduits are beneficial toregeneration. J Biomed Mater Res Part A 2007;80(4):965–82.

[28] Bian YZ, Wang Y, Aibaidoula G, Chen GQ, Wu Q. Evaluation of poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) conduits for peripheral nerveregeneration. Biomaterials 2009;30(2):217–25.

[29] Borschel GH, Kia KF, Kuzon WM, Dennis RG. Mechanical properties of acellularperipheral nerve. J Surg Res 2003;114(2):133–9.

[30] Ndreu A, Nikkola L, Ylikauppila H, Ashammakhi N, Hasirci V. Electropsun biode-gradable nanofibrous mats for tissue engineering. Nanomed 2008;3(1):45–60.

[31] Lu L, Peter SJ, Lyman MD, Lai HL, Leite SM, Tamada JA, et al. In vitro and in vivodegradation of porous poly(DL-lactic-co-glycolic acid) foams. Biomaterials2000;21(18):1837–45.

[32] Ferreira BMP, Zavaglia CAC, Duek EAR. Films of poly (L-lactic acid)/poly(hydroxybutyrate-co-hydroxyvalerate) blends: in vitro degradation. Mater Res2001;4(1):34–42.

[33] Moore MJ, Friedman JA, Lewellyn EB, Mantila SM, Krych AJ, Ameenuddin S,et al. Multiple-channel scaffolds to promote spinal cord axon regeneration.Biomaterials 2006;27(3):419–29.

[34] Rutkowski GE, Miller CA, Jeftinija S, Mallapragada SK. Synergistic effects ofmicropatterned biodegradable conduits and Schwann cells on sciatic nerveregeneration. J Neural Eng 2004;1(3):151–7.

[35] Corey JM, Lin DY, Mycek KB, Chen Q, Samuel S, Feldman EL, et al. Alignedelectrospun nanofibers specify the direction of dorsal root ganglia neuritegrowth. J Biomed Mater Res A 2007;83(3):636–45.

[36] Stokols S, Sakamoto J, Breckon C, Holt T, Weiss J, Tuszynski MH. Templatedagarose scaffolds support linear axonal regeneration. Tissue Eng2006;12(10):2777–87.

[37] Okamoto H, Hata KI, Kagami H, Okada K, Ito Y, Narita Y, et al. Recoveryprocess of sciatic nerve defect with novel bioabsorbable collagen tubespacked with filaments in dogs. J Biomed Mater Res Part A, in press, doi:10.1002/jbm.a.32421.