Introduction

Spinal cord injuries involving partial or complete tran-section, as with other lesions in the central nervoussystem, are unable to heal on their own. Complete

spinal cord injuries in humans and other mammalscause loss of sensory, motor and, reflex functions belowthe site of injury. Several different approaches have beenused in attempting to reconstruct an injured spinal cord.The use of growth factors, either by exogenous

S. Rochkind

A. Shahar

D. Fliss

D. El-Ani

L. Astachov

T. Hayon

M. Alon

R. Zamostiano

O. Ayalon

I. E. Biton

Y. Cohen

R. Halperin

D. Schneider

A. Oron

Z. Nevo

Development of a tissue-engineeredcomposite implant for treating traumaticparaplegia in rats

Received: 2 April 2004Revised: 23 April 2005Accepted: 6 June 2005Published online: 15 November 2005� Springer-Verlag 2005

Abstract This study was designed toassess a new composite implant toinduce regeneration of injured spinalcord in paraplegic rats followingcomplete cord transection. Neuronalxenogeneic cells from biopsies ofadult nasal olfactory mucosa(NOM) of human origin, or spinalcords of human embryos, were cul-tured in two consecutive stages: sta-tionary cultures in a viscous semi-solid gel (NVR-N-Gel) and in sus-pension on positively charged mi-crocarriers (MCs). A tissue-engineered tubular scaffold, con-taining bundles of parallel nanofi-bers, was developed. Both the tubeand the nanofibers were made of abiodegradable dextran sulphate–gel-atin co-precipitate. The suturablescaffold anchored the implant at thesite of injury and provided guidancefor the regenerating axons. Implantsof adult human NOM cells wereimplanted into eight rats, fromwhich a 4 mm segment of the spinalcord had been completely removed.Another four rats whose spinalcords had also been transected wereimplanted with a composite implant

of cultured human embryonic spinalcord cells. Eight other cord-tran-sected rats served as a control group.Physiological and behavioral analy-sis, performed 3 months afterimplantation, revealed partialrecovery of function in one or twolimbs in three out of eight animals ofthe NOM implanted group and in allthe four rats that were implantedwith cultured human embryonicspinal cord cells. Animals of thecontrol group remained completelyparalyzed and did not show trans-mission of stimuli to the brain. Theutilization of an innovative com-posite implant to bridge a gapresulting from the transection andremoval of a 4 mm spinal cord seg-ment shows promise, suggesting thefeasibility of this approach for par-tial reconstruction of spinal cord le-sions. Such an implant may serve asa vital bridging station in acute andchronic cases of paraplegia.

Keywords Olfactory mucosa ÆSpinal cord Æ Transection ÆTransplantation

Eur Spine J (2006) 15: 234–245DOI 10.1007/s00586-005-0981-8 ORIGINAL ARTICLE

S. Rochkind Æ A. Shahar (&)D. El-Ani Æ L. Astachov Æ T. HayonR. Zamostiano Æ O. Ayalon Æ Z. NevoNeural & Vascular Reconstruction Labs,Ness Ziona, IsraelE-mail: abrahams@nvrlab.comTel.: +972-8-9301099Fax: +972-8-9302662

S. Rochkind Æ D. Fliss Æ M. AlonDivision of Peripheral NerveReconstruction, Departments ofNeurosurgery and Otolaryngology,Tel Aviv Sourasky Medical Center,Tel Aviv University,Ramat Aviv, Tel-Aviv, Israel

I. E. Biton Æ Y. Cohen Æ Z. NevoDepartment of Clinical Biochemistry,The Raymond and Beverly SacklerFaculty of Exact Sciences,School of Chemistry,Sackler School of Medicine,Tel-Aviv University,Ramat Aviv, Tel-Aviv, Israel

R. Halperin Æ D. Schneider Æ A. OronDepartments of Gynecology,Obstetrics and Orthopedics,Assaf Harofeh Medical Center,Zerifin, Israel

administration [8, 14, 36] or by introducing growthfactor-treated implants [22, 23, 39] and geneticallyengineered cells [34, 46] have been attempted with lim-ited success. Others have concentrated their efforts onthe use of various tissue-engineered scaffolds [4, 54, 57].Spinal cord reconstruction using implantation of cellsfrom various sources has been gaining momentum inrecent years [28, 41]. However, one of the major disad-vantages of the implantation or injection of cells alone isthe limited viable cell survival after the procedure, ascells tend to desert the injury site.

One feasible innovative way of repairing injuredmammalian spinal cord is by creating a composite im-plant, which contains cultured cells from an autologousor allogeneic source. The implanted cells grow and serveas a vital bridge to connect the stumps of the severedspinal cords [44]. Such a composite spinal cord implantincludes a combination of the various elements thatsupport the healing process, such as a tissue-engineeredscaffold to anchor the implanted cells, which areembedded in a viscous adhesive milieu, supplementedwith a variety of factors [47].

The dream of using primordial cells from early humanembryos for implantation therapy in spinal cord injuriesis now a step closer to reality [27, 37]. Recent discoveriesextend the potential for isolation and cultivation of hu-man embryonic stem cells [49], and their differentiationin culture to neurons and other cell types [43, 49]. Fur-thermore, viable treatment of spinal cord injuries usingsuch human embryonic stem cells would be considered asan allogeneic graft, and should overcome the risks ofrejection, immunogenic reaction and possible neoplastictransformation.

An excellent autologous source of adult neuronalprecursor cells is the nasal olfactory mucosa (NOM).The NOM tissue is easily biopsied and the neurons andthe sustentacular cells of the NOM renew themselvesconstantly during life by proliferation of the basal globalstem cells [9, 15, 45]. Moreover, the unique sensorycharacteristics of the NOM neurons render them suit-able for receiving and transmitting signals as a ’’relaystation’’ in the spinal cord.

The scaffolds produced by tissue-engineering tech-nologies for spinal cord implantation are categorizedinto synthetic and naturally occurring polymeric mac-romolecules. The naturally occurring ones are commonlygrouped into protein-peptides-based scaffolds and car-bohydrates-based scaffolds [26, 29, 52]. There are severalpublished examples where the scaffold imitates theextracellular matrix (ECM) present in connective tissues,such as collagens and proteoglycans [55, 59, 60]. Thelatter type best matches the ECM constituents of manytissues, including that of the spinal cord.

Other promising strategies focus on creating areceptive environment for regeneration by abolishing themyelin-associated inhibitors of axonal growth and the

physical and biochemical barrier, the so-called ‘‘glialscar’’ [7, 17, 40, 48, 53].

Last, but not the least important, is the intercellularviscous adhesive milieu and its ingredients for support-ing the vitality of the neuronal bridge. The ECM groundsubstances of the spinal cord have the consistency ofcolloidal-gels, and indeed, several types of gels have beenproposed for repairing injured spinal cord [34, 56, 57].

In the current study we report on the development ofan innovative composite implant designed to serve as avital bridging station in acute and chronic cases of spinaltransection or injury.

Materials and methods

The composite implant elements

The scaffold is based on NVRs tissue-engineered tech-nology (patent pending), consisting of two naturallyoccurring biological cross-linked polymers. It has atubular format with a customized diameter of 2 mm anda wall thickness of 0.4 mm. In addition, the tubularscaffold contains a bundle of parallel nanofibers 50–100 lm in diameter, made of the same material as thescaffold (Fig. 1). In a series of preliminary experimentsthe scaffold was found to be biocompatible, non-toxic,and non-inflammatory (data not shown). The tube istransparent, suturable, and it can last for a period ofmore than 3 months until biodegradation takes place(data not shown).

NVR-N-Gel, which is a proprietary product ofNVR Labs (patent pending), is composed of a cross-linked hyaluronic acid with the adhesive moleculelaminin, and the following mixture of ingredients: an-tioxidants, neuronal growth factor; neuro-protective

Fig. 1 A tubular scaffold containing nanofibers. Original magnifi-cation 25 times

235

factors; EGF, bFGF, BDNF and NGF—20–50 ng/ml,IGF-I—50 ng/ml, LIF—0.5 u/ml, NAC (n-acetyl cy-stein)—10 lM, pifithrin a, cyclic—200 nM, and reti-noic acid—1–5 lM.

The NVR-N-Gel is transparent, highly hydrated withpolar and non-polar (hydrophobic) residues, all bio-compatible and biodegradable. For cell cultivation, thegel is used at a concentration of 0.7–1% hyaluronic acid,however, for implantation a more viscous gel (1.2–1.5%)is used.

Cultivation of adult human NOM and embryonic spinalcord cells

Biopsies of adult human olfactory nasal mucosa andembryonic spinal cords of aborted fetuses (16–23 weeksof gestation) were collected for cell isolation and theestablishment of cultures. The study with human derivedsamples was approved by the Helsinki committee ofAssaf-Harofeh Medical Center (no. 13/03). Informedconsent was obtained from each patient.

An innovative culture technique was introduced asoutlined in the scheme (Fig. 2). It combines stationarycultures in gel, alternating with cells grown in suspensionon an anion exchange, positively charged cylindrical,DEAE-cellulose (DE-53) microcarriers (MCs) (What-man, England). The MCs are equilibrated with phos-phate buffered saline (PBS) pH 7.4 and autoclaved inbatches of 15 g in 100 ml PBS [19, 51].

The dissociated cells were grown in suspension at-tached to the MCs for periods of 1–4 weeks. At varioustimes during their growth period in suspension, cell-MCaggregates were collected and reseeded in NVR-N-Gelas stationary cultures. The following growth media wereused: modified NEP medium based on the Bottensteinand Sato [6] N2 medium or M-21 medium. The NEPmedium contains DMEM-F12 (Invitrogen, UK), N2additives (progesterone, putresine, selenium, insulin andtransferrin) plus complex B27 supplements (Invitrogen)and 1% BSA. The M-21 medium is based on NEPmedium, except that B27 is omitted and 100 lM non-essential amino acids, 30 ng/ml triiodothyronine, and1 mM sodium pyruvate are added. Both media weresupplemented with the same factors used to enrich thegel. For cell expansion, the media were enriched with10% fetal calf serum (FCS); for growth differentiationand implantation the serum was omitted from themedia. Cells were usually cultured for 3–4 weeks prior toimplantation.

Surgical and transplantation procedure

The study was authorized by the local ethical committeefor experiments in laboratory animals. Animal treatmentand maintenance were in accordance with the ’’Guide forthe care and use of animals’’, Institute of laboratoryanimal resources commission on life sciences, NationalResearch Council, National Academy Press, Washing-ton, DC, 1966. The DHEW publication no. 80 (NIH)

Fig. 2 Preparation of adult hu-man NOM and human embry-onic SC cultures

236

Office of Science and Health reports DRR/NIH, Beth-esda, MD 20205, USA.

Twenty Sprague-Dawly rats, 3-months old, eachweighing approximately 250 g, were used. All rats wereanesthetized by intraperitoneal injections. For theoperations, anesthesia consisted of Ketamine HCL(50501 USA) 125–130 mg/kg and Xylosine (B2370 Bel-gium) 4.8 mg/kg. For electrophysiological tests, ratswere anesthetized for a short period (30 min) with Ke-tamine 75–85 mg/kg and Xylosine 3 mg/kg. For re-moval of stitches, light anesthesia was used: 50 mg/kgKetamine and 1.8 mg/kg Xylosine. All the treatmentsand the follow up tests were performed in a double blindrandomized manner.

All surgical procedures were performed on anesthe-tized rats, using a high magnification navigator micro-

scope (Zeiss NC-4) in a class 100 animal operatingroom. The spinal cord was exposed via a dorsal ap-proach. The overlying muscles were retracted, T7–T8laminae removed, the spinal cord was completely tran-sected using micro-scissors and a 4 mm segment of thecord was removed. A 4 mm gap was chosen to matchthe gap size formed after removal of the scar in chronicspinal cord injuries.

Eight of the 20 rats who underwent spinal cordtransection and the removal of a 4 mm spinal cordsegment, were closed with no further treatment (shamtreated control group). The remaining 12 rats underwentimplantation of one of the composite implants. Eightrats were treated with implants containing NOM cells(Fig. 3) and four rats received human embryonal spinalcord-derived cells.

Fig. 3 Cultured adult humanNOM neurons. a–d Sproutingof nerve fibers concomitantlywith migration of nerve cellsfrom M-cell aggregates inNVR-N-Gel. a–c—200 times,d—100 times. a–d Phase con-trast microscopy. Immunofluo-rescent staining of NOMneurons with antibodies specificto MAP 2 (e) and olfactorymucosa protein (OMP) (f).Original magnification e—400times, f—200 times, respectively

237

Embryonic human spinal cord cells were grown aslong term cultures to the stage of myelin formation(Fig. 4). About 1–2 · 106 NOM or spinal cord cells,embedded in NVR-N-Gel, and encapsulated in NVRscaffold, were implanted into the site of the excisedspinal cord segment.

Four-millimeter long composite implants were placedin the transected area of the spinal cord, in direct contactwith the margins of the two stumps. The entire area ofthe lesion containing the implant was covered with a thinbiodegradable membrane composed of the biological co-polymer, attached by a few interstitial sutures for fixa-tion of the implants at the desired sites. Finally, themuscular and cutaneous planes were closed and sutured.

Post-operative animal maintenance

In the post-operative management of the animals carewas taken to minimize discomfort and pain. Followingimplantation the rats were assisted in urination anddefecation with the help of a veterinarian, twice daily.Animals were maintained in ventilated cages, containingsterile sawdust and sterile food. The paraplegic rats werekept solitary in cages, but were gathered in groups for1 h every day, in a large facility. At the termination ofthe experiments, the animals were sacrificed under gen-eral anesthesia.

Electrophysiological measurements

Somatosensory evoked potentials (SSEP) were recordedin the experimental and control groups in a blindedmanner, immediately postoperatively and 3 months la-ter. Conductivity of the spinal cord was studied bystimulation of the sciatic nerve and recording from twodisc-recording electrodes, active and reference, placed onthe rats’ scalps. These electrodes, with conductive jelly,were attached to the scalp—active over the somatosen-sory cortex in the midline and reference electrode be-

tween the two eyes. The earth electrode was placed onthe thigh, on the side of the stimulation. The sciaticnerve was stimulated by a bipolar stimulating electrode.Two hundred and fifty-six stimulation pulses of 0.1 msin duration were generated at a rate of 3 s)1. The stim-ulus intensity was increased gradually, until slighttwitching of the limb appeared. The appearance ofevoked potentials, as a response to stimulation in twoconsecutive tests, was considered positive.

Latency and amplitude (positive—P wave peak) weremeasured

The rats were anesthetized intra-operatively with dilutedNembutal 15 mg/kg weight.

The test was performed using the Medelec/Teca Sap-phire 4 ME electromyography apparatus (20 Hz–2 KHzband pass filter and calibration sensitivity 10–20 mcV/divand time base 5 ms/div).

Locomotor rating scale

The BBB scale [5] describes the locomotor rating scale,graded from 0 (absent performance) to 21 (completenormal gait performance). This grading scale was usedto assess behavioral recovery and gait performanceweekly for 2–10 months after spinal cord injury.

Magnetic resonance imaging (MRI) analysis

Sample preparation: spinal cords were excised and fixedwith formalin. The spinal cords were inserted into 5 mmNMR tubes with their long axis parallel to the z-direction(the Bo direction) of the magnet and immersed in Flu-orinert (Sigma Chemical Co., St. Louis, MO, USA). Thetemperature in the magnet was maintained at25.0±0.1�C for the duration of the experiments.

Fig. 4 Phase contrast micros-copy of mature motor neuron ina and myelinated axons in b(arrows) in long-term culturesof human embryonic spinalcord cells. Original magnifica-tion 400 times

238

The MRI diffusion experiments were performed on awide-bore 8.4T NMR spectrometer (Bruker, Karlsruhe,Germany) equipped with a micro5 imaging gradientprobe capable of producing pulse gradients of up to190 G cm)1 in each of the three directions. Diffusionweighted MR images were collected using the stimulatedecho diffusion imaging pulse sequence with the followingparameters: TR=2000 ms, TE=35 ms, d=3 ms, and D=50 ms. The diffusion gradient strength, G, was incre-mented from 0 to 60 G cm)1 in 16 steps giving a maxi-mal b value of 1.12·106 s cm)2 and qmax of 766 cm)1.Diffusion was measured perpendicular and parallel tothe long axis of the spine. The MR images were collectedin a blind mode and the trauma site was placed at thecenter of the imaged region.

The signal decay of water was analyzed using the q-space approach [12] using the Matlab program.

Histological analyses

Prior to cultivation, a small segment of the NOM biopsywas taken for histological staining after fixation in 10%formalin at pH 7.4.

The spinal cord area, containing the implant regionwith both the proximal and the distal normal healthystumps, was fixed as a whole mount in 10% formalinfor several days. Subsequently, samples were placed inethylenediaminetetra-acetic acid (12.5%) at pH 7.0 fordecalcification. The softening of the calcified bonyspine enabled the smooth release of the whole region ofthe spinal cord, which could be inserted into the tubeprobe of the MRI system (described previously). Forhistological analysis, the spinal cord was dissected intoseven sections of 3–4 mm each, in parallel with theMRI analysis. The pieces were rinsed thoroughly inrunning tap water. The samples were dehydrated insequential alcohols, xylol and embedded in paraffin.Sections of 5–6 lm were re-hydrated and stained withMeyer’s hematoxylin-eosin, Masson’s trichrome, andBodian silver methods [42].

Immunofluorescence

Cultures were fixed with 4% paraformaldehyde andincubated with antibodies against epitopes of OlfactoryMarker Protein (OMP was kindly provided by Prof. F.Margolis, University of Maryland, Baltimore, MD,USA) and microtubulin associated protein-2 (MAP-2).Cells were then washed and incubated with CY2 orTexas Red conjugated goat anti-mouse or rabbit IgG(Jackson, diluted 1:100). The stained cultures were in-spected under fluorescent microscope (Olympus, NY,USA) with suitable filters and photographed usingMagnafire SP digital camera (Optronix, Mansfield NG,UK). For positive control, rat neuronal cultures werestained under the same conditions.

Results

When adult human NOM cells were seeded directly inNVR-N-Gel as primary stationary cultures, growth ofmorphologically different cell types was observed: epi-thelial cells organized in large laminae and tapered cellsforming islands of confluent cells, giving rise to largenumbers of round, unattached floating cell masses.

A large number of neuronal cells were observedamong the epithelial laminae and the tapered cells in thecultures that were reseeded in NVR-N-Gel, following agrowth period of 1–2 weeks in suspension on positivelycharged MCs (Fig. 4).

Post-operative observation

Seven animals out of eight in the control group exhib-ited complete paraplegic characteristics on physicalexamination and in their gait performance analysis(BBB=0: no observable hind limb movement). One ratof the control group showed slight movement of the hipjoints (BBB=1) (Table 1).

Table 1 BBB locomotor rating scale of the three groups of animals, two experimental and one control

Control Human nasal olfactory mucosa Human embryonic spinal cord

Rat No. Onset ofmovement(days)

BBBscale

Rat No. Start ofmovement(days)

BBB scale Rat No. Onset ofmovement(days)

BBB scale

1 – 0 1 15 10 1 15 132 – 0 2 – 0 2 15 63 – 0 3 13 13 3 48 104 – 0 4 92 3 4 12 35 – 0 5 – 06 – 0 6 – 07 30 1 7 – 08 – 0 8 – 0

239

Three of the eight rats in the group treated by acomposite implant containing cultured human NOMshowed varying degrees of active movements. One of thethree rats showed BBB=3 (extensive movement of hipand knee joints), one rat showed BBB=10 (occasionalweight-supported plantar steps; no FL–HL coordina-tion) and one rat showed BBB=13 (frequent consistentweight-supported plantar steps and FL–HL coordina-tion) (Table 1).

All four rats in the group treated by human embry-onic spinal cord cells implantation showed varying de-grees of leg movements: one rat showed BBB=3(extensive movement of hip and knee joints), one ratshowed BBB=6 (extensive movement of two joints andslight movement of the third), one rat showed BBB=10(occasional weight-supported plantar steps; no FL–HLcoordination), and one rat showed BBB=13 (frequentconsistent weight-supported plantar steps and FL–HLcoordination) (see Table 1, Fig. 5).

Statistical analyses of the three groups of animalsgraded by the BBB scale

The BBB scale does not follow a normal distribution(Shapiro-Wilk highly significant for two out of threegroups). A non-parametric test is therefore used tocompare the groups. The Kruskal–Wallis test shows asignificant difference between the groups (p=0.0117).

To test which of the group pairs is significantlydifferent, we used the Duncan multiple comparisontest on the ranked data (as the raw data is non-nor-mal). The mean rank of human embryonic spinal cordgroup (16.88) is significantly higher than both thehuman NOM and control groups (means of 10.5 and7.31, respectively). No significant difference emergedwhen comparing the NOM and control groups.

Electrophysiological measurements

Spinal cord conductivities were measured immediatelyafter spinal cord transection and again 3 months later intwo groups—transection alone, or transection plus

implantation of composite NOM implants. In two out ofthe three NOM implanted-rats exhibiting legs move-ments, SSEPs were elicited (Fig. 6b). No SSEP responsewas elicited in the eight rats of the control group, nor inthe five rats of the NOM implanted group that did notshow leg movement (Fig. 6a).

MRI analysis

Representatives of the treatment groups were subjectedto MRI analysis, which provided information on thestate of the spinal cord tissue at the injury site. Magneticresonance (MR) q-space displacement maps [1, 2], whichwere computed for three different spinal cords, revealedthat fiber-like tissue with an amount of water-restricteddiffusion was present only in the treated spinal cords andnot in the controls (Fig. 7). Moreover, comparison ofslices numbers 3–5, which are of the implantation sites,show small areas in which the mean displacement is lessthan 4 lm (value consistent with normal white matter).Such areas are present only in slices of implanted rats,but not in the controls (Fig. 7).

Histological analysis

The dominant histological picture of the reparative tis-sue in the area of the excised cord tissue was the presenceof fibrotic scar tissue, composed of glial cells and fi-broblasts, together with the formation of new bloodvessels. No inflammatory reaction was observed either inthe histology of rats implanted with human cells or inthe control rats.

In rats implanted with either human NOM or withembryonic spinal cord the H&E stained sections showedsome areas of neurokeratin (shrunk axons surroundedby an empty space, residual of the myelin sheath thathad been dissolved by the alcohol treatment). Further-more, in one rat (that was walking on the right leg afterNOM implantation) a number of large neuronal peri-karya were observed in the implanted area (Fig. 8a, c).In silver stained sections of both composite implants

Fig. 5 a Complete paralysis ofboth legs, folded inward, of acontrol rat that underwentcomplete transection of thespinal cord and removal of a4 mm segment. b Paraplegic ratshowing restoration of partialgait performance (in the rightleg) 3 weeks after implantationof a composite implant con-taining cultured adult humanNOM cells into a 4 mm gap oftransected spinal cord

240

several nerve fibers could be seen crossing the reparativetissue (Fig. 8b, d)

Discussion

Regeneration and repair of complete transection injuriesof the spinal cord is still an unresolved clinical challenge.Various experimental approaches for reconstructiveregeneration and renewal of damaged spinal cord areunder intensive investigation in laboratory animals. Thestrategies include stimulation of positive autoimmuneresponses [16, 50], introduction of neurotrophic andneuroprotective agents [8, 14, 22, 23, 34, 36], or removaland elimination of scar inhibitory molecules [11, 21, 24,30, 53]. The latest technologies for solving paraplegicconditions employ several kinds of cell therapy that areintroduced into the damaged site of the spinal cord afterremoval of the accumulated scar [47]:

(a) Implantation of tissue-engineered devices, withoutcells, for anchorage of the implant and for guidingaxonal regeneration [59, 60].

(b) Composite implants, containing cells of eitherautologous or allogeneic origin [18, 35, 37, 41]. Thechoice of cell resources involves implantation of ei-ther stem cells directed to differentiate toward ma-ture neurogenic phenotypes or insertion of alreadymature neuronal committed cells.

In the current project, both cell sources of the com-posite implants are of human origin, serving in the ratspinal cord system as xenogeneic cell implants; they aremeant in future clinical studies to simulate autologous(NOM cells) and allogeneic (embryonic spinal cord cells)sources, respectively. The innovative cell culture tech-niques described in the Methods section served thepurpose of establishing high concentrations of neuralcell cultures ideally [19, 51]. These cells are embedded ina milieu of a gel enriched with adhesive molecules [13,38] and neurotrophic and neuroprotective agents asantioxidants, which are released slowly. In addition, thegel embeds the nanofilaments in the tubular scaffold.The gel and the scaffold are biocompatible, bioerodable,and biodegradable.

Other reports in the literature describe auxiliary tis-sue-engineered products that promote neuronal andaxonal growth, such as nanofibrils [52, 59, 60], gels [34,56, 57] and scaffolds [22, 54, 55]. Acellular devices yiel-

Fig. 6 a Absence of spinal cord conductivity (SSEP) in aparaplegic control rat after complete transection of the spinal cordand removal of 4 mm segment. b Restoration of spinal cordconductivity after complete transection and implantation ofcomposite implant containing NOM

Fig. 7 q-Space displacement maps (x direction) sequential slices ofMRI analyses

241

ded poorer and limited results in comparison to cell-containing implants [47].

A variety of sources for implantation have been tes-ted, including bone marrow stroma cells [25, 33, 58],skin, umbilical cord blood and embryonal and fetal stemcell lines [20, 37, 41, 43, 61]. However, those cell sourceshave proven to support only sporadic appearances ofsingle cells or cell aggregates of neuronal cells, and failedto yield robust numbers of neuronal cells to create asuccessful implant that can replace massive segmentallosses. In the current study, embryonic spinal cord andadult NOM cells showed vigorous vitality that estab-lished rich neurogenic cell cultures for implantation.

Both types of xenogeneic implants used in the currentstudy bear the risk of inducing an immunogenic response.However, surprisingly, the results show that neither ofthem evoked a rejection response. On the other hand,

preliminary results of studies currently in progress showthat when cultured NOM cells of adult dogs were im-planted into paralytic rats, severe rejection led to the deathof animals a few days after the implantation. It is impor-tant to note that the spinal cord and its surrounding fluidsare considered an immunologically privileged site, withminimal immunologic responses.

The human embryonal spinal cord implants in ourstudy showed significant advantages over the controlanimals, although repair of the tissue damage was by nomeans complete. The rats implanted with the adult hu-man NOM cells showed statistically insignificantadvantages over the control group of animals. In thecourse of further experiments being performed; we aregaining more insight into the growth of the human adultNOM biopsy, by further selecting and characterizingspecific sub-populations of cells from the overall col-lected mixture. The outcome of these experiments mayfurther clarify the contribution of the different adulthuman NOM cell-populations.

Over the recent years MRI has developed into themost important imaging modality for the central nervoussystem (CNS). Water diffusion is an important contrastmechanism in MRI of the CNS, diffusion weightedimaging (DWI) and diffusion tensor imagings (DTI) are

Fig. 8 Histological sections of implanted spinal cords 10 months(a–c) after adult NOM implantation and 3 months (d) afterimplantation of human embryonic spinal cord cells. Hematoxy-lin-eosin (H&E) staining demonstrates dispersed neuronal peri-karya (a, arrows). Silver staining demonstrates nerve fibers—eithersingle (b, arrows), or organized in parallel bundles (d, arrows). Inaddition, note areas of neurokeratin (c, arrows). Original magni-fication 400 times

242

used extensively for studying CNS pathologies andstructures [31, 32]. The DTI is currently also used tostudy fiber orientations [31]. Recently, we have demon-strated that high b-value q-space diffusion MRI is ex-tremely sensitive to the structural and the patho-physiological state of axons and fibers of the CNS [1–3,10], and thus can be used to study CNS maturation anddegeneration [3]. Interestingly, in the current study thefirst two rats transplanted with cell-containing compositeimplants showed some regeneration of motor functionand axons, consistent with the electrophysiologicalmeasurements, MRI, and some histological evidence offormation of cells and axonal fibers. It is clear thatmore MRI experiments are needed to verify if thesepreliminary results imply that axonal connection wasindeed established across the transplanted zone. SuchMRI experiments are underway.

The MRI analyses are supported by both the elec-trophysiological tests and the histological findings ofneuronal cells and crossing axons along sections of thecomposite implanted transplants, versus their absence inthe sections of the control animals.

In summary, in the current study we describe thecombined efforts of cell cultivation, differentiation andintegration into a tissue-engineered scaffold, as a com-posite implant for repairing spinal cord injury. Reliabletests to evaluate the final outcomes of the spinal cordinjured regions yielded results that brought us closer tothe goal of treatment of damaged spinal cord. Never-theless, we did not face the problem of elimination ofscar tissue, which still needs to be addressed. These re-sults are encouraging, but we have still a long way to gobefore complete motor function and full gait perfor-mance are restored.

References

1. Assaf Y, Cohen Y (2000) Assignment ofthe water slow-diffusing component inthe central nervous system using q-space diffusion MRS: implications forfiber tract imaging. Magn Reson Med43:191–199

2. Assaf Y, Mayk A, Cohen Y (2000)Displacement imaging of spinal cordusing q-space diffusion-weighted MRI.Magn Reson Med 44:713–722

3. Assaf Y, Mayk A, Eliash S, Speiser Z,Cohen Y (2003) Hypertension andneuronal degeneration in excised ratspinal cord studied by high-b value q-space diffusion magnetic resonanceimaging. Exp Neurol 184:726–736

4. Bakshi A, Fisher O, Dagci T, HimesBT, Fischer I, Lowman A (2004)Mechanically engineered hydrogel scaf-folds for axonal growth and angiogen-esis after transplantation in spinal cordinjury. J Neurosurg Spine 1:322–329

5. Basso DM, Beattie MS, Bresnahan JC(1995) A sensitive and reliable locomo-tor rating scale for open field testing inrats. J Neurotrauma 12:1–21

6. Bottenstein JE, Sato GH (1979) Growthof a rat neuroblastoma cell line in ser-um-free supplemented medium. ProcNatl Acad Sci USA 76:514–517

7. Bradbury EJ, Moon LD, Popat RJ,King VR, Bennett GS, Patel PN,Fawcett JW, McMahon SB (2002)Chondroitinase ABC promotes func-tional recovery after spinal cord injury.Nature 416:636–640

8. Bregman BS, McAtee M, Dai HN,Kuhn PL (1997) Neurotrophic factorsincrease axonal growth after spinal cordinjury and transplantation in the adultrat. Exp Neurol 148:475–494

9. Calof AL, Bonnin A, Crocker C, Ka-wauchi S, Murray RC, Shou J, Wu HH(2002) Progenitor cells of the olfactoryreceptor neuron lineage. Microsc ResTech 58:176–188

10. Cohen Y, Assaf Y (2002) High b-valueq-space analyzed diffusion-weightedMRS and MRI in neuronal tissues—atechnical review. NMR Biomed 15:516–542

11. Condic ML, Lemons ML (2002)Extracellular matrix in spinal cordregeneration: getting beyond attractionand inhibition. Neuroreport 13:A37–A48

12. Cory DG, Garroway AN (1990) Mea-surement of translational displacementprobabilities by NMR: an indicator ofcompartmentation. Magn Reson Med14:435–444

13. Costa S, Planchenault T, Charriere-Bertrand C, Mouchel Y, Fages C, Juli-ano S, Lefrancois T, Barlovatz-MeimonG, Tardy M (2002) Astroglial permis-sivity for neuritic outgrowth in neuron-astrocyte cocultures depends on regu-lation of laminin bioavailability. Glia37:105–113

14. Coumans JV, Lin TT, Dai HN, Mac-Arthur L, McAtee M, Nash C, Breg-man BS (2001) Axonal regeneration andfunctional recovery after completespinal cord transection in rats by de-layed treatment with transplants andneurotrophins. J Neurosci 21:9334–9344

15. Cunningham AM, Manis PB, Reed RR,Ronnett GV (1999) Olfactory receptorneurons exist as distinct subclasses ofimmature and mature cells in primaryculture. Neuroscience 93:1301–1312

16. David S, Ousman SS (2002) Recruitingthe immune response to promote axonregeneration in the injured spinal cord.Neuroscientist 8:33–41

17. Fouad K, Dietz V, Schwab ME (2001)Improving axonal growth and func-tional recovery after experimental spinalcord injury by neutralizing myelinassociated inhibitors. Brain Res BrainRes Rev 36:204–212

18. Galvin KA, Jones DG (2002) Adulthuman neural stem cells for cell-replacement therapies in the centralnervous system. Med J Aust 177:316–318

19. Goldman SA, Nedergaard M, CrystalRG, Fraser RA, Goodman R, Harri-son-Restelli C, Jiang J, Keyoung HM,Leventhal C, Pincus DW, Shahar A,Wang S (1997) Neural precursors andneuronal production in the adult mam-malian forebrain. Ann NY Acad Sci835:30–55

243

20. Gorio A, Torrente Y, Madaschi L, DiStefano AB, Pisati F, Marchesi C, Bel-icchi M, Di Giulio AM, Bresolin N(2004) Fate of autologous dermal stemcells transplanted into the spinal cordafter traumatic injury (TSCI). Neuro-science 125:179–189

21. Hermanns S, Klapka N, Muller HW(2001) The collagenous lesion scar—anobstacle for axonal regeneration inbrain and spinal cord injury. RestorNeurol Neurosci 19:139–148

22. Houle JD, Ziegler MK (1994) Bridginga complete transection lesion of adultrat spinal cord with growth factor-treated nitrocellulose implants. J NeuralTransplant Plast 5:115–124

23. Houweling DA, Lankhorst AJ, GispenWH, Bar PR, Joosten EA (1998) Col-lagen containing neurotrophin-3 (NT-3)attracts regrowing injured corticospinalaxons in the adult rat spinal cord andpromotes partial functional recovery.Exp Neurol 153:49–59

24. Jones LL, Yamaguchi Y, Stallcup WB,Tuszynski MH (2002) NG2 is a majorchondroitin sulfate proteoglycan pro-duced after spinal cord injury and isexpressed by macrophages and oligo-dendrocyte progenitors. J Neurosci22:2792–2803

25. Kabos P, Ehtesham M, Kabosova A,Black KL, Yu JS (2002) Generation ofneural progenitor cells from whole adultbone marrow. Exp Neurol 178:288–293

26. Kataoka K, Suzuki Y, Kitada M, Ha-shimoto T, Chou H, Bai H, Ohta M,Wu S, Suzuki K, Ide C (2004) Alginateenhances elongation of early regenerat-ing axons in spinal cord of young rats.Tissue Eng 10:493–504

27. Keirstead HS (2001) Stem cell trans-plantation into the central nervous sys-tem and the control of differentiation. JNeurosci Res 63:233–236

28. Keyvan-Fouladi N, Raisman G, Li Y(2003) Functional repair of the corti-cospinal tract by delayed transplanta-tion of olfactory ensheathing cells inadult rats. J Neurosci 23:9428–9434

29. Kisiday J, Jin M, Kurz B, Hung H,Semino C, Zhang S, Grodzinsky AJ(2002) Self-assembling peptide hydrogelfosters chondrocyte extracellular matrixproduction and cell division: implica-tions for cartilage tissue repair. ProcNatl Acad Sci USA 99:9996–10001

30. Krautstrunk M, Scholtes F, Martin D,Schoenen J, Schmitt AB, Plate D,Nacimiento W, Noth J, Brook GA(2002) Increased expression of theputative axon growth-repulsive extra-cellular matrix molecule, keratan sul-phate proteoglycan, following traumaticinjury of the adult rat spinal cord. ActaNeuropathol (Berl) 104:592–600

31. Le Bihan D (2003) Looking into thefunctional architecture of the brain withdiffusion MRI. Nat Rev Neurosci4:469–480

32. Le Bihan D, van Zijl P (2002) Fromthe diffusion coefficient to thediffusion tensor. NMR Biomed 15:431–434

33. Lee JB, Kuroda S, Shichinohe H, YanoS, Kobayashi H, Hida K, Iwasaki Y(2004) A pre-clinical assessment modelof rat autogeneic bone marrow stromalcell transplantation into the centralnervous system. Brain Res Brain ResProtoc 14:37–44

34. Loh NK, Woerly S, Bunt SM, WiltonSD, Harvey AR (2001) The regrowth ofaxons within tissue defects in the CNS ispromoted by implanted hydrogelmatrices that contain BDNF andCNTF producing fibroblasts. ExpNeurol 170:72–84

35. Lu J, Feron F, Mackay-Sim A, WaitePM (2002) Olfactory ensheathing cellspromote locomotor recovery after de-layed transplantation into transectedspinal cord. Brain 125:14–21

36. Lu P, Yang H, Jones LL, Filbin MT,Tuszynski MH (2004) Combinatorialtherapy with neurotrophins and cAMPpromotes axonal regeneration beyondsites of spinal cord injury. J Neurosci24:6402–6409

37. McDonald JW, Liu XZ, Qu Y, Liu S,Mickey SK, Turetsky D, Gottlieb DI,Choi DW (1999) Transplanted embry-onic stem cells survive, differentiate andpromote recovery in injured rat spinalcord. Nat Med 5:1410–1412

38. McKeon RJ, Hoke A, Silver J (1995)Injury-induced proteoglycans inhibitthe potential for laminin-mediated axongrowth on astrocytic scars. Exp Neurol136:32–43

39. Meijs MF, Timmers L, Pearse DD,Tresco PA, Bates ML, Joosten EA,Bunge MB, Oudega M (2004) Basicfibroblast growth factor promotes neu-ronal survival but not behavioralrecovery in the transected and Schwanncell implanted rat thoracic spinal cord. JNeurotrauma 21:1415–1430

40. Moon LD, Asher RA, Rhodes KE,Fawcett JW (2001) Regeneration ofCNS axons back to their target follow-ing treatment of adult rat brain withchondroitinase ABC. Nat Neurosci4:465–466

41. Ogawa Y, Sawamoto K, Miyata T,Miyao S, Watanabe M, Nakamura M,Bregman BS, Koike M, Uchiyama Y,Toyama Y, Okano H (2002) Trans-plantation of in vitro-expanded fetalneural progenitor cells results in neuro-genesis and functional recovery afterspinal cord contusion injury in adultrats. J Neurosci Res 69:925–933

42. Pearse A (1972) Histochemistry theo-retical and applied, vols 1&2, 3rd edn.Little Brown and Co., Boston, MA

43. Reubinoff BE, Pera MF, Fong CY,Trounson A, Bongso A (2000) Embry-onic stem cell lines from human blast-ocysts: somatic differentiation in vitro.Nat Biotechnol 18:399–404

44. Rochkind S, Shahar A, Amon M, NevoZ (2002) Transplantation of embryonalspinal cord nerve cells cultured on bio-degradable microcarriers followed bylow power laser irradiation for thetreatment of traumatic paraplegia inrats. Neurol Res 24:355–360

45. Roisen FJ, Klueber KM, Lu CL,Hatcher LM, Dozier A, Shields CB,Maguire S (2001) Adult human olfac-tory stem cells. Brain Res 890:11–22

46. Ruitenberg MJ, Levison DB, Lee SV,Verhaagen J, Harvey AR, Plant GW(2005) NT-3 expression from engineeredolfactory ensheathing glia promotesspinal sparing and regeneration. Brain128:839–853

47. Schmidt CE, Leach JB (2003) Neuraltissue-engineering: strategies for repairand regeneration. Annu Rev BiomedEng 5:293–347

48. Schnell L, Schwab ME (1990) Axonalregeneration in the rat spinal cord pro-duced by an antibody against myelin-associated neurite growth inhibitors.Nature 343:269–272

49. Schuldiner M, Eiges R, Eden A, Ya-nuka O, Itskovitz-Eldor J, GoldsteinRS, Benvenisty N (2001) Induced neu-ronal differentiation of human embry-onic stem cells. Brain Res 913:201–205

50. Schwartz M (2003) Sell memorial lec-ture. Helping the body to cure itself:immune modulation by therapeuticvaccination for spinal cord injury.J Spinal Cord Med 26(Suppl 1):S6–S10

51. Shahar A (1990) Cultivation of nerveand muscle cells on microcarriers.Methods Neurosci 2:195–209

52. Silva GA, Czeisler C, Niece KL, Beni-ash E, Harrington DA, Kessler JA,Stupp SI (2004) Selective differentiationof neural progenitor cells by high-epi-tope density nanofibers. Science303:1352–1355

53. Silver J, Miller JH (2004) Regenerationbeyond the glial scar. Nat Rev Neurosci5:146–156

54. Stokols S, Tuszynski MH (2004) Thefabrication and characterization of lin-early oriented nerve guidance scaffoldsfor spinal cord injury. Biomaterials25:5839–5846

244

55. Tatard VM, Menei P, Benoit JP, Mon-tero-Menei CN (2005) Combiningpolymeric devices and stem cells for thetreatment of neurological disorders: apromising therapeutic approach. CurrDrug Targets 6:81–96

56. Woerly S, Doan VD, Evans-Martin F,Paramore CG, Peduzzi JD (2001) Spinalcord reconstruction using NeuroGelimplants and functional recovery afterchronic injury. J Neurosci Res 66:1187–1197

57. Woerly S, Petrov P, Sykova E, RoitbakT, Simonova Z, Harvey AR (1999)Neural tissue formation within poroushydrogels implanted in brain and spinalcord lesions: ultrastructural, immuno-histochemical, and diffusion studies.Tissue Eng 5:467–488

58. Woodbury D, Schwarz EJ, Prockop DJ,Black IB (2000) Adult rat and humanbone marrow stromal cells differentiateinto neurons. J Neurosci Res 61:364–370

59. Yoshii S, Oka M (2001) Collagen fila-ments as a scaffold for nerve regenera-tion. J Biomed Mater Res 56:400–405

60. Yoshii S, Oka M, Shima M, Akagi M,Taniguchi A (2003) Bridging a spinalcord defect using collagen filament.Spine 28:2346–2351

61. Zigova T, Song S, Willing AE, HudsonJE, Newman MB, Saporta S, Sanchez-Ramos J, Sanberg PR (2002) Humanumbilical cord blood cells express neu-ral antigens after transplantation intothe developing rat brain. Cell Trans-plant 11:265–274

245