Hindawi Publishing CorporationMediators of In!ammationVolume "#$%, Article ID %"&'(', $" pageshttp://dx.doi.org/$#.$$))/"#$%/%"&'('

Research ArticlePalmitoylethanolamide Is a Disease-Modifying Agent inPeripheral Neuropathy: Pain Relief and Neuroprotection Sharea PPAR-Alpha-Mediated Mechanism

L. Di Cesare Mannelli,1 G. D’Agostino,2 A. Pacini,3 R. Russo,2 M. Zanardelli,1

C. Ghelardini,1 and A. Calignano2

! Department of Preclinical and Clinical Pharmacology, University of Florence, Viale Pieraccini ", #$!%& Florence, Italy'Department of Experimental Pharmacology, University of Naples Federico II, ($!%! Naples, Italy%Department of Anatomy, Histology and Forensic Medicine, University of Florence, #$!%& Florence, Italy

Correspondence should be addressed to L. Di Cesare Mannelli; lorenzo.mannelli@uni*.it

Received + November "#$"; Revised "' December "#$"; Accepted " January "#$%

Academic Editor: Kuen-Jer Tsai

Copyright © "#$% L. Di Cesare Mannelli et al. ,is is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Neuropathic syndromes which are evoked by lesions to the peripheral or central nervous system are extremely di-cult to treat,and available drugs rarely joint an antihyperalgesic with a neurorestorative e.ect. N-Palmitoylethanolamine (PEA) exerts anti-nociceptive e.ects in several animalmodels and inhibits peripheral in!ammation in rodents. Aimed to evaluate the antineuropathicproperties of PEA, a damage of the sciatic nerve was induced in mice by chronic constriction injury (CCI) and a subcutaneousdaily treatment with %#mgkg!1 PEA was performed. On the day $/, PEA prevented pain threshold alterations. Histological studieshighlighted that CCI induced oedema and an important in*ltrate of CD&+ positive cells in the sciatic nerve. Moreover, osmicatedpreparations revealed a decrease in axon diameter and myelin thickness. Repeated treatments with PEA reduced the presence ofoedema andmacrophage in*ltrate, and a signi*cant highermyelin sheath, axonal diameter, and a number of *bers were observable.In PPAR-! null mice PEA treatment failed to induce pain relief as well as to rescue the peripheral nerve from in!ammation andstructural derangement.,ese results strongly suggest that PEA, via a PPAR-!-mediated mechanism, can directly intervene in thenervous tissue alterations responsible for pain, starting to prevent macrophage in*ltration.

1. Introduction

Neuropathic painmay originate from several di.erent causes.Mechanical peripheral neuropathies are the consequence oflocal or extrinsic compression phenomena or impingementby an anatomic neighbor causing a localized entrapment.Traumatic neuropathies are the result of either closed injuriesor open injuries to peripheral nerves [$, "]. Both of thesecategories are characterized by an important in!ammatorycomponent that plays a central role in the pathogenesis ofneuropathic pain. In!ammatory cells (e.g.,macrophages), theproduction of molecules that mediate in!ammation (cyto-kines), and the production of nervous growth factors areinvolved [%, /]. In animal models it has been demonstratedthat peripheral nerve injuries induce a profound local in!am-matory response that involves T cells andmacrophages [)]. In

particular, in the neuropathic painmodel induced by chronicconstriction injury (CCI) an important macrophage in*ltratehas been described in the damaged sciatic nerve [)–'] andin the dorsal root ganglia [&, (]. ,e in!ammatory responseparalleled with nervous tissue alterations and pain [', $#].

N-Palmitoylethanolamine (PEA), the endogenous amidebetween palmitic acid and ethanolamine, belongs to the fam-ily of fatty acid ethanolamides (FAEs), a class of lipid medi-ators. PEA exerts antinociceptive e.ects in several animalmodels [$$, $"], prevents neurotoxicity and neurodegenera-tion [$%, $/], and inhibits peripheral in!ammation and mastcell degranulation [$)]. Endogenous and exogenous PEA canmodulate macrophage response [$+, $'].

Anti-in!ammatory e.ects of PEA have been associatedwith peroxisome proliferator-activated receptor-(PPAR-)!activation [$&], a nuclear receptor fundamental in the control

" Mediators of In!ammation

of in!ammatory responses, and expressed in various cellsof the immune system [$(, "#]. PEA does not elicit anti-in!ammatory e.ects in mutant PPAR-! null mice (PPAR-!!/!). Indeed, when assessed in either the carrageen hindpawor phorbol ester ear pinna tests, PEA reduced in!ammationin wild-type, but not in PPAR-!!/!, mice [$&]. On the otherhand, LoVerme and collegues [$"] demonstrated the pivotalrole of PPAR-! in the PEA pharmacodynamic mechanism torelieve pain.

In a mouse peripheral neuropathy model (CCI) we eval-uated the e.ects of repeated PEA treatments on the sciaticnerve lesions responsible for neuropathic pain. Aimed tohighlight the role of PPAR-! in PEA-evoked neurorestorationduring neuropathy, a morphological study has been perfor-med in both wild-type and PPAR-! null mice.

2. Materials and Methods

'.!. Animals. All procedures met the European guidelinesfor the care and use of laboratory animals (&+/+#(/ECCand "#$#/+%/UE), and those of the Italian Ministry ofHealth (DL $$+/("). Male wild-type (WT) and PPAR-!!/! (KO) (B+.$"(S/-SvJae-Pparatm$Gonz) mice, previouslybackcrossed to C)'BL+mice for $# generations, were bred inour animal facility, where a colony was established andmain-tained by heterozygous crossing. Mice were genotyped asdescribed on the supplier webpage (http://jaxmice.jax.org/),with minor modi*cations. DNA was extracted from tailsusing the RedExtract kit (Sigma-Aldrich, Milan, Italy). Allanimals were maintained on a $" h light/$" h dark cycle withfree access to water and standard laboratory chow.

'.'. Chemicals. PEA was from Tocris (Bristol, UK); it wasdissolved in PEG andTween &# " : $ (Sigma-Aldrich) and keptovernight under gentle agitation with a microstirring bar.Before injection, sterile saline was added so that the *nalconcentrations of PEG and Tween &# were "# and $#%v/v, respectively. Drug was injected subcutaneously (s.c.) ina dose of %#mgkg!1—#.%mL mouse—for consecutively $/days from the day a0er surgery.

'.%. CCI Model. ,e sciatic nerve of )-+-week-old WT andKO mice were surgically ligated as described ["$]. In brief,the animals were anesthetized with ketamine ($##mgkg!1i.p) and xylazine ($#mgkg!1 i.p.), the le0 thigh was shavedand scrubbed with Betadine, and a small incision (" cm inlength) was made in the middle of the le0 thigh to exposethe sciatic nerve.,e nerve was loosely ligated at two distinctsites (spaced at a "mm interval) around the entire diameterof the nerve using silk sutures ('-#). Behavioral tests wereperformed on the day $/ a0er surgery.

'.). Mechanical Hyperalgesia. We measured mechanicalhyperalgesia using a Randall-Selitto analgesimeter for mouse(Ugo Basile, Varese, Italy). Latencies of paw withdrawal to acalibrated pressure were assessed on ipsilateral (ligated) pawson day $/ following ligatures. Cut-o. force was set at +# g.

'.#. Mechanical Allodynia. To assess for changes in sensationor in the development of mechanical allodynia, sensitivity totactile stimulation was measured using the Dynamic Plan-tar Aesthesiometer (DPA, Ugo Basile, Italy). Animals wereplaced in a chamber with a mesh metal !oor covered bya plastic dome that enabled the animal to walk freely butnot to jump. ,e mechanical stimulus was then deliveredin the mid-plantar skin of the hind paw. ,e cuto. was*xed at ) g. Each paw was tested twice per session. ,is testdid not require any special pretraining, just an acclimationperiod to the environment and testing procedure. Testingwasperformed on ipsilateral (ligated) paw on day $/ a0er ligation.Cuto. force was set at ) g.

'.". Tissue Explants. A0er the algesic test, animals weresacri*ced and the ipsilateral sciatic nerves, $ cm proximal anddistal to the ligation, were explanted; the portion containingthe ligature was eliminated. Contralateral nerves were alsodissected, and an equivalent portion was collected. Spinalcord was removed, and lumbar section was immediatelyfrozen in liquid nitrogen.

'.*. Osmic Acid Staining. ,e sciatic nerve was stored in a /%glutaraldehyde solution. ,e tissue samples were osmicatedin $% solution of osmium tetroxide for " h under constantagitation. Before and a0er osmication, the tissue was repeat-edly rinsed in #.$M sodium cacodylate at pH './. A0ergradual dehydration in ethanol, the osmicated nerve sampleswere embedded in para-n (Diapath,Milan, Italy). Transverse) "m sections were cut on a Reichert microtome (Leica,Rijswijk, ,e Netherlands), mounted with Canada balsam,and observed under a light microscopy.

'.(. Azan-Mallory Stain. A0er the sacri*ce, the sciatic nervewas *xed in situ using /% formalin in phosphate bu.er (pH'./), nerves were *xed in /% bu.ered neutral formalin solu-tion, and then the nerve was embedded in para-n. Finally) "m sections were stained with Azan-Mallory for lightmicroscopy studies andwere graded for oedema and in*ltrate[""]. ,e sections were semiquanti*ed by an arbitrary scalestarting from $, mild in*ltrate and oedema, up to $#, severein*ltrate and widespread oedema.,e procedure was carriedout by an independent researcher who was masked to theexperiment.

'.&. Morphometry. Morphometry was performed on crosssections of osmium-*xed sciatic nerves $#mm starting fromthe level of injury or at the corresponding level in uninjuredcontrol nerves. ,e $#mm cross section corresponded to alevel distal or proximal to the injury. Counts and measure-ments were carried out using ImageJ analysis so0ware.

,e *rst step of morphometric analysis of the sciaticnerve consisted of identifying and capturing the entire fasci-cle image, followed by measurement of the fascicle perimeterand area by contouring its internal epineural (magni*cation:objective "#x). ,e next step consisted of capturing sequen-tial inner areas of the fascicle (magni*cation: objective $##x).,e high-magni*cationmicrographs were randomly selected

Mediators of In!ammation %

0

1

2

3

4

5 Von Frey

VehiclePEA

Paw

with

draw

al th

resh

old

(g)

!!

+/+ "/"30mg kg"1(a)

0

10

20

30

40

50

60 Randall

Paw

with

draw

al th

resh

old

(s)

!!

VehiclePEA

+/+ "/"30mg kg"1(b)

F12345 $: Chronic treatment e.ects on pain behaviour. Comparison of PEA-e.ects in PPAR-!+/+ and !/! mice. (a) Response to a mechanicalnoxious stimulus on the ipsilateral paw $/ days a0er CCI evaluated by Randall-Selitto test. (b) Response to amechanical nonnoxious stimulusevaluated by Von Frey test. PEA (%#mgkg!1) was daily s.c. administered for $/ days starting from the day of operation. Each value representsthe mean of " experiments with $"mice per group. ""# < 0.01 versus vehicle-treated mice.

in nonoverlapping areas to cover )#–')% of the total cross-sectional area of the nerve.,e random selected histologicalimages were converted into binary (black and white) imagesand cleaned of any blood vessels, degenerated nerve *bers,and artifacts, and the following parameters were then mea-sured: total number and density of nerve *bers, axon dia-meter, and myelin sheath area. ,e number of small *bers,de*ned as *bers < / "m, and large *bers, de*ned as *bers$ / "m, was calculated, and the myelin thickness was deter-mined from the di.erence in perimeter between the *ber andthe axon. ,e procedure was carried out by an independentresearcher who was masked to the experiment.

'.!$. Immunohistochemistry. Sections of ) "m were deparaf-*nized, dehydrated, and submitted to antigen retrieval. Sec-tions were then blocked by incubation with %% normal goatserum for "#min. Macrophages were detected using anti-CD&+ antibody. ,e slides were then incubated with theprimary antibodies for $& h at /#C. A0er washing in TBS, thesectionswere treatedwith the secondary antibody conjugatedto Alexa Fluor /&& ($ : $###, Invitrogen, Milan, USA) for $ hat room temperature. Nuclei were counterstained with DAPI.Quantitative analysis of CD&+ positive cells was performedcollecting at least three independent *elds through a 20 %0.5NA objective, and positive cells were counted using the“cell counter” plugin of ImageJ.

'.!!. Western Blot Analysis. ,e lumbar portion of the spinalwas homogenized in lysis bu.er containing )#mM Tris-HCl pH &.#, $)#mM NaCl, $mM EDTA, #.)% Triton X-$##,

Complete Protease Inhibitor (Roche), and the homogenatewas incubated on ice for %# minutes. ,en, the suspensionwas sonicated on ice using three $#-second bursts at highintensity with a $#-second cooling period between eachburst. A0er centrifugation ($%### %g for $) minutes at /#C)aliquots containing "#"g total protein underwent to westernblot analysis using a mouse anti-COX" antiserum ($ : $###;Cell Signalling, USA). Densitometric analysis was performedusing the “ImageJ” analysis so0ware, and results were nor-malized to &-actin immunoreactivity ($ : $### rabbit anti-serum, SantaCruz Biotechnology) as internal control.

'.!'. Statistical Analysis. For behavioral experiments, resultswere expressed as the mean ± SEM, and analyses were con-ducted using Statistica (Statso0, Tulsa, OK, USA).

Histological, morphometric, and immunohistochemicalanalyses were performed on ) mice per group, evaluating+ sciatic nerve sections for each animal. Comparison, werecarried out using Mann-Whitney nonparametric tests. In allcases, the investigator was blind to the experimental statusof each animal. Slides from control and experimental groupswere labeled with numbers so that the person performingthe image analysis was blinded as to the experimental group.In addition, all images were captured and analyzed by aninvestigator other than the one who performed measures toavoid possible bias. Western blot analysis were performedon ) mice per group; comparisons were carried out usingBonferroni posttest. Data were analyzed using the “Origin'.)” so0ware. Di.erences were considered signi*cant if # <0.05 or otherwise di.erently reported.

/ Mediators of In!ammation

50

75

100

125

150

175

200

225N

umbe

r of !

bers

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"1

!!#

(a)

50

75

100

125

150

175

200

225

Num

ber o

f !be

rs

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

30mg kg"1

! !

Prox.Dist.

(b)

F12345 ": Morphometry: number of *bers. )"m nerve sections of osmium-*xed tissues were analyzed.,e distal and the proximal tract ofthe ipsilateral ligated nerve (CCI) was compared with the contralateral and with the sciatic nerve of sham-operated animals (sham). E.ectof repeated PEA administrations (%#mgkg!1 s.c. for $/ days) on the number of nervous *bers in respect to vehicle-treated CCI animals andvehicle-treated sham animals. (a) PPAR-!+/+ mice versus (b) knock-out animals. Quantitative analysis was performed evaluating ) animalsfor each group. "# < 0.05 was considered as signi*cantly di.erent from sham, vehicle-treated mice. ## < 0.05 was considered as signi*cantlydi.erent from CCI, vehicle-treated mice.

3. Results

PEA, (%#mgkg!1) s.c. daily administered starting on the dayof operation, prevented pain threshold alterations elicited byCCI (Figure $). In wild-type animals $/ days a0er injury, PEAreduced the hypersensitivity to a mechanical noxious stimu-lus (Randall-Selitto test; Figure $(a)) as well as the hypersen-sitivity to a nonnoxious mechanical stimulus (Figure $(b)).PEA e-cacy against CCI-evoked pain was lacked in PPAR-! knock-out ('/') mice (Figures $(a) and $(b)).

On the $/th day a0er injury, a morphological evaluationof the sciatic nerves was performed both on the proximaland distal parts from the ligation. )"m sections of para-n-embedded nerve were stained by the Azan-Mallory pro-cedure revealing that CCI was able to induce a strongeralteration in the distal portion of the ipsilateral nerve. ,ehistological analysis highlighted an extensive demyelination,myelin abnormalities like characteristic aggregate, generallytermed “ovoids”, pathognomonic of myelin degeneration(Figure ', arrows), and an axonal damage. As shown inFigure " the number of *bers was signi*cantly reduced,particularly in the distal part of the nerve of both wild-type(+/+; graph (a)) and PPAR-! null ('/'; graph (b)) animals.

Morphometric evaluation of osmium-*xed tissuesallowed the characterization of the alteration of the myelinsheet thickness and axonal diameter, as well as the discrimi-nation of large against small *bers (strati*ed by diameterin >+"m for large and <+"m for small). CCI was able todecrease the myelin thickness of large and small *bers in thedistal portion of the ipsilateral nerve in respect to the sham

both in wild-type and in knock-out mice (Figures %(a) and%(b), large *bers; Figures %(c) and %(d), small *bers).

In regard to axon diameter, a time-dependent decreasewas revealed for all the *bers, particularly the small type, bothin the distal and in the proximal portions of the ipsilateralnerve; morphometry revealed a similar pro*le in PPAR-!+/+(Figures /(a) and /(c)) and PPAR-!!/! (Figures /(b) and/(d)) animals. In wild-type mice repeated PEA administra-tions, %#mg kg!1 for $/ days, were able to preserve the nervemorphology. Nerve sections of PEA-treated mice showed ahigher number of*bers in respect to the saline-treated groups(Figure "(a)); the myelin thickness in the distal portions ofthe nerve was decreased to a lesser extent (Figures %(a) and%(c)); the axon diameter was protected in the PEA group bothin the proximal and in the distal nerve parts, even in the small*bers (Figures /(a) and /(c)). PEA was completely ine.ectivein PPAR-! null mice in preventing sciatic nerve alterationsevaluated as number of *bers (Figure "(b)), myelin thickness(Figures %(b) and /(d)) and axon diameters (Figures /(b) and/(d)).

Azan-Mallory staining revealed an abundant in!amma-tory in*ltrate in the ligated nerve. Figure ) shows the in*ltrateevaluation $/ days a0er ligation: in!ammatory cells werepresent in the proximal and, at higher level, in the distalparts of both PPAR-!+/+ (graph (a)) and PPAR-!!/! (graph(b)) mice. %#mgkg!1PEA signi*cantly prevented the cellularin*ltrate number in wild-type (Figure )(a)) but not in KOanimals (Figure )(b)).

Mediators of In!ammation )

1

1.25

1.5

1.75M

yelin

thick

ness

($m)Large !bers

!!#

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

Prox.Dist.

30mg kg"1(a)

1

1.25

1.5

1.75

Mye

lin th

ickne

ss ($m)

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"1

Large !bers

! !

(b)

0.25

0.5

0.75

1

1.25

Mye

lin th

ickne

ss ($m)

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"1

Small !bers

!!#

(c)

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"1

Mye

lin th

ickne

ss ($m)

0.25

0.5

0.75

1

1.25Small !bers

! !

(d)

F12345 %: Morphometry: myelin thickness. Nerve sections ()"m) of osmium-*xed tissues were analyzed.,e distal and the proximal tract ofthe ipsilateral ligated nerve (CCI) was compared with the contralateral and with the sciatic nerve of sham-operated animals (sham). Myelinthickness of large and small *bers of PEA-treated (%#mg kg!1 s.c. for $/ days) PPAR-!+/+ ((a) and (c)) and PPAR-!!/! ((b) and (d)) micein respect to saline-treated CCI and saline-treated sham animals. Quantitative analysis was performed evaluating ) animals for each group."# < 0.05 was considered as signi*cantly di.erent from sham, vehicle-treated mice. ## < 0.05 was considered as signi*cantly di.erent fromCCI, vehicle-treated mice.

Moreover, both osmium *xed and Azan-Mallory-stainedsections (Figure ') allowed the observation of amassive pres-ence of oedema among the *bers of CCI animals. Figure +show the quantitative oedema evaluation $/ days a0er oper-ation: the alteration was more evident in the distal portionthan in the proximal one without revealable di.erences dueto knock down PPAR-! gene. PEA (%#mgkg!1 s.c. for $/

days) was able to prevent the oedema induction of about )#%in CCI wild-type mice (Figure +(a)). No oedema protectivee.ects were observable in PEA-treated PPAR-!!/! animals inrespect to vehicle (Figure +(b)).

,e immune in!ammatory cells evaluated were dif-fusely distributed throughout the nerve tissue in all samplesof the CCI mice, whereas a mild CD&+ positive reaction

+ Mediators of In!ammation

1

2

3

4

5

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

Prox.Dist.

30mg kg"1

Large !bersAx

on d

iamet

er ($m) !#

!!!

#

(a)

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

Prox.Dist.

30mg kg"1

Large !bers

1

2

3

4

5

Axon

diam

eter

($m)

!!!!

(b)

1

1.5

2

2.5

3

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

Prox.Dist.

30mg kg"1

Small !bers

Axon

diam

eter

($m) !#!#

!!

(c)

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral

Prox.Dist.

30mg kg"1

Small !bers

1

1.5

2

2.5

3Ax

on d

iamet

er ($m)

!

!

!!

(d)

F12345 /: Morphometry: axon diameters. Nerve sections ()"m) of osmium-*xed tissues were analyzed.,e distal and the proximal tract ofthe ipsilateral ligated nerve (CCI) was compared with the contralateral and with the sciatic nerve of sham-operated animals (sham). Axondiameters of large and small *bers of PEA-treated (%#mg kg!1 s.c. for $/ days) wild-type ((a) and (c)) and PPAR-! null ((b) and (d)) micein respect to saline-treated CCI and saline-treated sham animals. Quantitative analysis was performed evaluating ) animals for each group."# < 0.05 was considered as signi*cantly di.erent from sham, vehicle-treated mice. ## < 0.05 was considered as signi*cantly di.erent fromCCI, vehicle-treated mice.

was detectable in CCI mice administered with PEA aswell as in sham-operated animals. PPAR-!!/! mice showeda persistence of macrophage in*ltrate also in the nerve ofPEA-treated animals (Figures & and ().

CCI-dependent in!ammatory response in the centralnervous system was evaluated measuring COX" levels. Asshown in Figure $#, in wild-type mice PEA was able to sig-ni*cantly prevent COX" increase induced by nerve ligation.,is anti-in!ammatory e.ect was lost in PPAR-!!/! animals.

4. Discussion

Pharmacological treatment of peripheral neuropathy is actu-ally restricted to symptomatic drugs that are only partiallyable to control pain perception ["%]. Antidepressants, antin-covulsants, and opioids cannot intervene in nervous tissuealterations that act as a base of neuropathic pain.,e presentresults describe the direct protective e.ect of repeated PEAtreatment on lesioned peripheral nerves.

Mediators of In!ammation '

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"10123456789

10In!l

trate

(a.u

.) !#!

!!

#In!ltrate

(a)

VehicleVehicle PEA30mg kg"1

0123456789

10

In!l

trate

(a.u

.)

!!

!!

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

In!ltrate

(b)

F12345 ): In*ltrate evaluation. On the $/th day a0er CCI, ) "mnerve sections of formalin-*xed tissues were analyzed by Azan-Mallory stain,and the presence of in!ammatory in*ltrate was evaluated and quanti*ed by an arbitrary scale starting from $, mild in*ltrate, up to $#, severein*ltrate. E.ect of PEA repeated treatments (%#mgkg!1 s.c. daily) was observed in (a) PPAR-!+/+ and in (b) !/! mice where the distal andthe proximal tract of the ipsilateral ligated nerve of CCI was compared with the contralateral and with the sciatic nerve of sham-operatedanimals. Quantitative analysis was performed evaluating ) animals for each group. "# < 0.05 was considered as signi*cantly di.erent fromsham, vehicle-treated mice. ## < 0.05 was considered as signi*cantly di.erent from CCI, vehicle-treated mice.

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

30mg kg"10123456789

10

Oed

ema (

a.u.)

Oedema

!#!

!!

#

(a)

Vehicle

CCI, contralateralSham CCI, ipsilateral Prox.

Dist.

0123456789

10

Oed

ema (

a.u.)

Oedema

!!!!

Vehicle PEA30mg kg"1(b)

F12345 +: Oedema evaluation. On the $/th day a0er CCI, ) "m nerve sections of formalin-*xed tissues were analyzed by Azan-Mallory stainand the presence of oedema in*ltrate was evaluated and quanti*ed by an arbitrary scale starting from $, mild oedema, up to $#, widespreadoedema. E.ect of PEA repeated treatments (%#mgkg!1 s.c. daily) was observed in (a) PPAR-!+/+ and in (b) !/! mice where the distal andproximal tract of the ipsilateral ligated nerve of CCI was compared with the contralateral and with the sciatic nerve of sham-operated animals.Quantitative analysis was performed evaluating ) animals for each group. "# < 0.05 was considered as signi*cantly di.erent from sham,vehicle-treated mice. ## < 0.05 was considered as signi*cantly di.erent from CCI, vehicle-treated mice.

CCI induces morphometric alterations of the sciaticnerve that dramatically a.ect the portion distal from theinjury and that are also able to induce severe proximal impair-ment. At the same dose active in pain relief, PEA preventsthe reduction of myelin sheet thickness and axonal diameterand improves a characteristic degeneration of myelin as

highlighted by Azan-Mallory staining. According to previousresults [', "/, ")] CCI-mediated nerve architecture derange-ment is accompanied by a profound local in!ammatory reac-tion which includes oedema, in*ltration of hematogenousimmune cells, and induction of various soluble factors likecytokines and chemokines. In particular, the present results

& Mediators of In!ammation

Sham CCI CCI + PEA

+/+

20x

"/"F12345 ': Light micrographs from ) "m transverse sections of mouse sciatic nerve stained by Azan-Mallory, comparison between PPAR-!+/+and PPAR-!!/! animals; $/th day. Sham: section of sciatic nerve from sham animals; CCI + vehicle: distal part of the sciatic nerve of injuredvehicle-treated animals; CCI + PEA distal part of the sciatic nerve of CCI mice s.c. treated with %#mgkg!1 PEA daily for $/ days startingfrom the surgery. Ovoids are evidenced by arrows. Original magni*cation "#x.

0

10

20

30

40

50

60

70

80

90

VehicleVehicle PEA

CCI, contralateralSham CCI, ipsilateral Dist.

30mg kg"1

Num

ber C

D86

pos

itive

cells

!!

#

(a)

VehicleVehicle PEA30mg kg"1

0

10

20

30

40

50

60

70

80

90

Num

ber C

D86

pos

itive

cells

! !

CCI, contralateralSham CCI, ipsilateral Dist.

(b)

F12345 &: CD&+ positive cells evaluation in sciatic nerve. $/ days a0er CCI, )"m sections of the formalin-*xed distal part of the sciatic nerveunderwent immunohistochemical staining for CD&+. E.ect of PEA repeated treatments (%#mgkg!1 s.c. daily) was evaluated in (a) PPAR-!+/+ and in (b) PPAR-!!/! mice, and quantitative analysis was performed evaluating ) animals for each group. "# < 0.05 was considered assigni*cantly di.erent from sham, vehicle-treated mice. ## < 0.05 was considered as signi*cantly di.erent from CCI, vehicle-treated mice.

describe a characteristic in*ltrate of CD&+ positive cells.CD&+ is a phenotypic marker of the “classically activated”M$ macrophages stimulated by proin!ammatory cytokines,as IFN(, or by lipopolysaccharide and typically recruited a0ernervous system trauma ["+]. M$ macrophages produce highlevels of oxidative metabolites (e.g., nitric oxide and superox-ide) and proin!ammatory cytokines that are essential for hostdefense and tumor cell killing but that also cause collateraltissue damage ["']. ,e treatment with PEA attenuates thedegree of peripheral in!ammation, reducing oedema and

macrophage in*ltration allowing for hypothesizing a syner-gism between the anti-in!ammatory and the neuroprotectivemechanisms of PEA. On the other hand, an in!ammationcontrol mediated by PEA is highlighted also in the spinalcord. According to previous data CCI induces a COX-"increase in locations of the central nervous system consistentwith the neuroanatomical substrates of spinal nociception["&]. COX-" produces prostaglandins that contribute to thedevelopment and maintenance of spinal hyperexcitabilitya0er peripheral nerve injury ["&, "(]. Reducing COX-" levels,

Mediators of In!ammation (

Sham CCI

+/+

+/+

+/+

"/"

CCI + PEA CCI + PEA

F12345 (: CD&+ positive cells evaluation in sciatic nerve. $/ days a0er CCI, )"m sections of the formalin-*xed distal part of the sciatic nerveunderwent immunohistochemical staining for CD&+. E.ect of PEA repeated treatments (%#mgkg!1 s.c. daily) was evaluated in PPAR-!+/+and in !/! mice, and representative images are showed, and a detailed image is shown in the insert. Original magni*cation "#x.

PEA seems able to intervene also in these centralmechanismsof pain chronicization. To note that a direct inhibition ofpro-in!ammatory cytokines, using a TNF-! antibody, forinstance, attenuates pain-related behavior but has no e.ecton nerve regeneration [%#].

PEA is a naturally occurring amide between palmitic acidand ethanolamine it is a lipid messenger known to mimicseveral endocannabinoid-driven actions even though PEAdoes not bind CB$, CB", and abn-CBD receptors [%$]. So far,numerous actions of PEA on immune cells such as inhibitionof mast cell degranulation, attenuation of leukocyte extrava-sation, andmodulation of cytokine release frommacrophageshave been described [$+, %"]. Anti-in!ammatory e.ects ofPEA have been associated with peroxisome proliferator-activated receptor (PPAR)-alpha activation [$&]. PPAR-!,well known for its role in lipid metabolism, controls tran-scriptional programs involved in the development of in!am-mation through mechanisms that include direct interactionswith the proin!ammatory transcription factors, NF-)B andAP$, and modulation of IkB function [%%]. Pharmacologicalstudies have demonstrated that PPAR-! agonists are ther-apeutically e.ective in rodent models of in!ammatory andautoimmune diseases [%/]. Our results show that in a neuro-pathic pain model the PPAR-! genetic ablation determinesa loss of PEA e.ectiveness in reducing oedema prevention

and CD-&+ positive in*ltrating cells. On the other hand, therecruitment of reactive in!ammatory cells and subsequentproin!ammatory cascades o.ers a prime target for neuropro-tective agents. Agonists of PPAR-! such as feno*brate andWy-$/+/% protect against cerebral injury by antioxidant andanti-in!ammatory mechanisms and reduce the incidence ofstroke in mice [%), %+]. Using a spinal cord injury model,Genovese et al. [%'] demonstrated that dexamethasone uti-lizes PPAR-! to reduce in!ammation and tissue injury in arat model of spinal cord trauma. On the contrary, the PPAR-! agonist gem*brozil does not promote tissue preservationand behavioral recovery a0er spinal contusion injury in mice[%&]. Our study shows the obligatory role of PPAR-! for theneuroprotective e.ect of PEA in peripheral neuropathy. Inthe sciatic nerve of CCI mice PEA exerts a widespread pro-tective e.ect on both myelin, and axons throughout a PPAR-!-mediated mechanism, since PEA treatment fails to rescuenerve tissue in PPAR-! knock-out animals. ,e neuropro-tective PEA properties were suggested by Skaper et al. [%(]since dose dependently PEA protected cerebellar granulecells from glutamate toxicity in neuronal single cell culturesand prevented histamine-induced cell death in hippocampalcultures. ,ese e.ects were elicited without involvement ofCB receptors. More recently Koch et al. [/#] described theprotective e.ect of PEA on dentate gyrus granule cells

$# Mediators of In!ammation

0

0.5

1

1.5

2

Vehicle Vehicle PEA30mg kg"1PEA30mg kg"1PPAR- PPAR-

O.D

. (de

nsito

met

ric u

nits/

mm2 )

COX-2

%-actin

!!&"/"&+/+

F12345 $#: COX-" expression levels in spinal cord. On the $/th day the spinal cord of PPAR-!+/+ and PPAR-!!/! mice was analyzed bywestern blot. Upper panel shows the densitometric quanti*cation of COX-" levels in CCI vehicle-treated animal in comparison with CCImice repetitively treated with PEA (%#mgkg!1 s.c. daily). Values were normalized to &-actin immunoreactivity. Data are expressed as themean ± SEM of triplicate determinations performed on ) animals for each group. ""# < 0.01 was considered as signi*cantly di.erent fromvehicle (+/+).

in excitotoxically lesioned organotypic hippocampal slicecultures; the speci*c PPAR-! antagonist GW+/'$ blockedthese e.ects. PEA exerts neuroprotective activities in neu-rodegenerative diseases. In amousemodel of Alzheimer’s andParkinson’s diseases, PEA reduced oxidative and apoptoticdamages and improve behavioral dysfunctions by a PPAR-!-mediated mechanism [$/, /$]. In a cellular model, PEA wasable to blunt &-amyloid-induced astrocyte activation and,subsequently, to improve neuronal survival through selectivePPAR-! activation [/"].

In neuropathic conditions PPAR-! seems to join the anti-hyperalgesic, anti-in!ammatory, and neuroprotective e.ectsof PEA. On the other hand the in!ammatory response toa damage is crucial for both pain sensation and tissue alte-rations; the importance of in!ammatory mediators is welldemonstrated in the pathogenesis of neuropathic pain, wherein*ltrating macrophages and Schwann cells may be involvedin the modulation of these mediators in response to nerveinjury [/%].

Starting from the relevance of the PPAR-! in PEA anti-neuropathic properties, the misunderstood role of perox-isome is intriguing. Peroxisomes ful*ll multiple tasks inmetabolism and adapt contents and functions according tocell type, age, and organism. Among the metabolic reac-tions that take place in peroxisomes, oxygen metabolism,&-oxidation of a number of carboxylates that cannot be

handled bymitochondria, !-oxidation of %-methyl-branchedchain and "-hydroxy fatty acids, ether lipid synthesis, anddetoxi*cation of glyoxylate are the most important [//,/)]. Patients with peroxisomal dysfunction present withsevere and diverse neurological anomalies, including neu-ronal migration defects, dysmyelination and in!ammatorydemyelination, macrophage in*ltration, and axon damage,proving that these organelles are indispensible for the normaldevelopment and maintenance of the nervous system [/)–/'].,erea0er, the peroxisome stimulation could be a broadspectrum approach to prevent nervous tissue damage, anda PEA, PPAR-!-mediated, increase in peroxisome numberand/or functionality could be also hypothesized. A preclinicalstudy [/&] showed that PEA-mediated reduction of spinalcord damage was paralleled by an induction of PPAR-!expression and an up-regulation of potential PPAR-! targetgenes, but a clear relationship betweenPPAR-! activation andperoxisome boosting is still lacking.

5. Conclusions

,e present results demonstrate the neuroprotective proper-ties of PEA in a preclinical model of neuropathic pain. Anti-hyperalgesic and neuroprotective properties are related to theanti-in!ammatory e.ect of PEA and its ability to prevent

Mediators of In!ammation $$

macrophage in*ltration in the nerve. PPAR-! stimulation isthe common pharmacodynamic code.

Acknowledgments

,is work was supported by the Italian Ministry of Instruc-tion, University and Research. ,e authors declare that theyhave no *nancial or other relationships that might lead to acon!ict of interests.

References

[$] P. J.Dyck, “,e causes, classi*cation and treatment of peripheralneuropathy,” +e New England Journal of Medicine, vol. %#', pp."&%–"&+, $(&".

["] R. Baron, “Mechanisms of disease: neuropathic pain—a clinicalperspective,” Nature Clinical Practice Neurology, vol. ", pp. ()–$#+, "##+.

[%] M. Tal, “A role for in!ammation in chronic pain,” CurrentReview of Pain, vol. %, pp. //#–//+, $(((.

[/] E. Vivoli, L. Di Cesare Mannelli, A. Salvicchi et al., “Acetyl-l-carnitine increases artemin level and prevents neurotrophicfactor alterations during neuropathy,”Neuroscience, vol. $+', no./, pp. $$+&–$$'/, "#$#.

[)] N. Uceyler, K. Gobel, S. G. Meuth et al., “De*ciency of thenegative immune regulator B'-H$ enhances in!ammation andneuropathic pain a0er chronic constriction injury of mousesciatic nerve,” Experimental Neurology, vol. """, no. $, pp. $)%–$+#, "#$#.

[+] D. Gomez-Nicola, B. Valle-Argos, M. Suardiaz, J. S. Taylor, andM. Nieto-Sampedro, “Role of IL-$) in spinal cord and sciaticnerve a0er chronic constriction injuty: regulation of macro-phage and T-cell in*latration,” Journal of Neurochemistry, vol.$#', pp. $'/$–$')", "##&.

['] A. Pacini, L. Di Cesare Mannelli, L. Bonaccini, S. Ronzoni,A. Bartolini, and C. Ghelardini, “Protective e.ect of alpha'nAChR: behavioural and morphological features on neuropa-thy,” Pain, vol. $)#, no. %, pp. )/"–)/(, "#$#.

[&] P. Hu and E. M. McLachlan, “Macrophage and lymphocyteinvasion of dorsal root ganglia a0er peripheral nerve lesions inthe rat,” Neuroscience, vol. $$", no. $, pp. "%–%&, "##".

[(] N.Morin, S. A. Owolabi,M.W.Harty et al., “Neutrophils invadelumbar dorsal root ganglia a0er chronic constriction injury ofthe sciatic nerve,” Journal of Neuroimmunology, vol. $&/, no. $-",pp. $+/–$'$, "##'.

[$#] L. Di Cesare Mannelli, C. Ghelardini, M. Calvani et al., “Pro-tective e.ect of acetyl-L-carnitine on the apoptotic pathway ofperipheral neuropathy,” European Journal of Neuroscience, vol."+, no. /, pp. &"#–&"', "##'.

[$$] A. Calignano, G. La Rana, A. Giu.rida, and D. Piomelli, “Con-trol of pain initiation by endogenous cannabinoids,”Nature, vol.%(/, no. ++(#, pp. "''–"&$, $((&.

[$"] J. LoVerme, R. Russo, G. La Rana et al., “Rapid broad-spec-trum analgesia through activation of peroxisome proliferator-activated receptor-!,” Journal of Pharmacology and Experimen-tal +erapeutics, vol. %$(, no. %, pp. $#)$–$#+$, "##+.

[$%] D. M. Lambert, S. Vandevoorde, G. Diependaele, S. J. Gov-aerts, and A. R. Robert, “Anticonvulsant activity of N-palmi-toylethanolamide, a putative endocannabinoid, inmice,”Epilep-sia, vol. /", no. %, pp. %"$–%"', "##$.

[$/] G. D’Agostino, R. Russo, C. Avagliano, C. Cristiano, R. Meli,and A. Calignano, “Palmitoylethanolamide protects against theamyloid-&")-%)-induced learning and memory impairment inmice, an experimental model of Alzheimer disease,” Neuropsy-chopharmacology, vol. %', pp. $'&/–$'(", "#$".

[$)] S. Mazzari, R. Canella, L. Petrelli, G. Marcolongo, and A.Leon, “N-("-Hydroxyethyl) hexadecanamide is orally active inreducing edema formation and in!ammatory hyperalgesia bydown-modulating mast cell activation,” European Journal ofPharmacology, vol. %##, no. %, pp. ""'–"%+, $((+.

[$+] R.A. Ross,H.C. Brockie, andR.G. Pertwee, “Inhibition of nitricoxide production in RAW"+/.' macrophages by cannabinoidsand palmitoylethanolamide,” European Journal of Pharmacol-ogy, vol. /#$, no. ", pp. $"$–$%#, "###.

[$'] C. Solorzano, C. Zhu, N. Battista et al., “Selective N-acyl-ethanolamine-hydrolyzing acid amidase inhibition reveals a keyrole for endogenous palmitoylethanolamide in in!ammation,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. $#+, no. /(, pp. "#(++–"#('$, "##(.

[$&] J. Lo Verme, J. Fu, G. Astarita et al., “,e nuclear receptorperoxisome proliferator-activated receptor-!mediates the anti-in!ammatory actions of palmitoylethanolamide,” MolecularPharmacology, vol. +', no. $, pp. $)–$(, "##).

[$(] R. A. Daynes and D. C. Jones, “Emerging roles of PPARs inin!ammation and immunity,”Nature Reviews Immunology, vol.", no. $#, pp. '/&–')(, "##".

["#] J. LoVerme,G. LaRana, R. Russo,A.Calignano, andD. Piomelli,“,e search for the palmitoylethanolamide receptor,” Life Sci-ences, vol. '', no. $/, pp. $+&)–$+(&, "##).

["$] R. Russo, J. LoVerme, G. La Rana et al., “,e fatty acidamide hydrolase inhibitor URB)(' (cyclohexylcarbamic acid%*-carbamoylbiphenyl-%-yl ester) reduces neuropathic paina0er oral administration in mice,” Journal of Pharmacology andExperimental +erapeutics, vol. %"", no. $, pp. "%+–"/", "##'.

[""] M. Kihara, P. J. Zollman, J. D. Schmelzer, and P. A. Low, “,ein!uence of dose of microspheres on nerve blood !ow, elec-trophysiology, and *ber degeneration of rat peripheral nerve,”Muscle and Nerve, vol. $+, no. $", pp. $%&%–$%&(, $((%.

["%] N. Attal, G. Cruccu, R. Baron et al., “EFNS guidelines on thepharmacological treatment of neuropathic pain: "#$# revision,”European Journal of Neurology, vol. $', article $$$%–e&&, "#$#.

["/] J. G. Cui, S. Holmin, T. Mathiesen, B. A. Meyerson, and B.Linderoth, “Possible role of in!ammatory mediators in tactilehypersensitivity in rat models of mononeuropathy,” Pain, vol.&&, no. %, pp. "%(–"/&, "###.

[")] C. Kleinschnitz, J. Brinkho., C. Sommer, and G. Stoll, “Con-tralateral cytokine gene induction a0er peripheral nerve lesions:dependence on the mode of injury and NMDA receptorsignaling,”Molecular Brain Research, vol. $%+, no. $-", pp. "%–"&,"##).

["+] K. A. Kigerl, J. C. Gensel, D. P. Ankeny, J. K. Alexander, D. J.Donnelly, and P. G. Popovich, “Identi*cation of two distinctmacrophage subsets with divergent e.ects causing either neu-rotoxicity or regeneration in the injured mouse spinal cord,”Journal of Neuroscience, vol. "(, no. /%, pp. $%/%)–$%///, "##(.

["'] A. H. Ding, C. F. Nathan, and D. J. Stuehr, “Release of reac-tive nitrogen intermediates and reactive oxygen intermediatesfrom mouse peritoneal macrophages: comparison of activatingcytokines and evidence for independent production,” Journal ofImmunology, vol. $/$, no. ', pp. "/#'–"/$", $(&&.

["&] H. L. Willingale, N. J. Gardiner, N. McLymont, S. Giblett,and B. D. Grubb, “Prostanoids synthesized by cyclo-oxygenase

$" Mediators of In!ammation

isoforms in rat spinal cord and their contribution to the devel-opment of neuronal hyperexcitability,” British Journal of Phar-macology, vol. $"", no. &, pp. $)(%–$+#/, $(('.

["(] X. Y. Zhu, C. S. Huang, Q. Li et al., “p%## exerts an epigeneticrole in chronic neuropathic pain through its acetyltransferaseactivity in rats following chronic constriction injury (CCI),”Molecular Pain, vol. &, p. &/, "#$".

[%#] T. Lindenlaub, P. Teuteberg, T. Hartung, and C. Sommer,“E.ects of neutralizing antibodies toTNF-alpha on pain-relatedbehavior and nerve regeneration in mice with chronic cons-triction injury,”Brain Research, vol. &++, no. $-", pp. $)–"", "###.

[%$] K. Mackie and N. Stella, “Cannabinoid receptors and endo-cannabinoids: evidence for new players,”+e AAPS Journal, vol.&, pp. E"(&–E%#+, "##+.

[%"] E. V. Berdyshev, “Cannabinoid receptors and the regulation ofimmune response,” Chem. Phys. Lipids, vol. $#&, pp. $+(–$(#,"###.

[%%] C. K. Glass and S. Ogawa, “Combinatorial roles of nuclearreceptors in in!ammation and immunity,” Nature ReviewsImmunology, vol. +, pp. //–)), "##+.

[%/] D. S. Straus and C. K. Glass, “Anti-in!ammatory actions ofPPAR ligands: new insights on cellular and molecular mecha-nisms,” Trends in Immunology, vol. "&, pp. ))$–))&, "##'.

[%)] H. Inoue, X. F. Jiang, T. Katayama, S. Osada, K. Umesono,and S. Namura, “Brain protection by resveratrol and feno*brateagainst stroke requires peroxisome proliferator-activated recep-tor ! in mice,”Neuroscience Letters, vol. %)", no. %, pp. "#%–"#+,"##%.

[%+] D. Deplanque, P. Gele, O. Petrault et al., “Peroxisome proli-ferator-activated receptor-! activation as a mechanism of pre-ventive neuroprotection induced by chronic feno*brate treat-ment,” Journal of Neuroscience, vol. "%, no. $), pp. +"+/–+"'$,"##%.

[%'] T. Genovese, E. Esposito, E. Mazzon et al., “PPAR-! modulatethe anti-in!ammatory e.ect of glucocorticoids in the secondarydamage in experimental spinal cord trauma,” PharmacologicalResearch, vol. )(, no. ), pp. %%&–%)#, "##(.

[%&] A. Almad,A. T. Lash, P.Wei, andA. E. Lovett-Racke, “,e PPARalpha agonist gem*brozil is an ine.ective treatment for spinalcord injured mice,” Experimental Neurology, vol. "%", pp. %#(–%$', "#$$.

[%(] S. D. Skaper, A. Buriani, R. Dal Toso et al., “,e ALIAmidepalmitoylethanolamide and cannabinoids, but not anandamide,are protective in a delayed postglutamate paradigm of excito-toxic death in cerebellar granule neurons,” Proceedings of theNational Academy of Sciences of the United States of America,vol. (%, no. (, pp. %(&/–%(&(, $((+.

[/#] M. Koch, S. Kreutz, C. Bottger et al., “Palmitoylethanolamideprotects dentate gyrus granule cells via peroxisome proliferator-activated receptor-alpha,” Neurotoxicity Research, vol. $(, no. ",pp. %%#–%/#, "#$$.

[/$] E. Esposito, D. Impellizzeri, E. Mazzon, I. Paterniti, and S. Cuz-zocrea, “Neuroprotective activities of Palmitoylethanolamidein an animal model of parkinson’s disease,” PLoS ONE, vol. ',article e/$&&#, "#$".

[/"] C. Scuderi, M. Valenza, C. Stecca, G. Esposito, M. R. Carratu,and L. Steardo, “Palmitoylethanolamide exerts neuroprotectivee.ects inmixed neuroglial cultures and organotypic hippocam-pal slices via peroxisome proliferator-activated receptor-alpha,”Journal of Neuroin,ammation, vol. (m article /(, "#$".

[/%] P. L. Woodhams, R. E. MacDonald, S. D. Collins, I. P. Chessell,and N. C. Day, “Localisation and modulation of prostanoid

receptors EP$ and EP/ in the rat chronic constriction injurymodel of neuropathic pain,” European Journal of Pain, vol. $$,no. +, pp. +#)–+$%, "##'.

[//] G. P. Mannaerts and P. P. Van Veldhoven, “Metabolic role ofmammalian peroxisomes,” in Proceedings of the Peroxisomes:Biology and Importance in Toxicology and Medicine, G. Gibsonand B. Lake, Eds., pp. $(–+", Taylor & Francis, London, UK,$((%.

[/)] R. J. Wanders and H. R. Waterham, “Peroxisomal disorders:the single peroxisomal enzyme de*ciencies,” in Biochimica etBiophysica Acta, vol. $'+%, pp. $'#'–$'"#, "##+.

[/+] S. Steinberg, G. Dodt, G. V. Raymond, N. E. Braverman, A. B.Moser, and H. W. Moser, “Peroxisome biogenesis disorders,”Biochimica et Biophysica Acta, vol. $'+%, pp. $'%%–$'/&, "##+.

[/'] M. Baes and P. Aubourg, “Peroxisomes,myelination, and axonalintegrity in the CNS,” Neuroscientist, vol. $), pp. %+'–%'(, "##(.

[/&] T. Genovese, E. Esposito, E. Mazzon et al., “E.ects of palmi-toylethanolamide on signaling pathways implicated in the deve-lopment of spinal cord injury,” Journal of Pharmacology andExperimental +erapeutics, vol. %"+, no. $, pp. $"–"%, "##&.