upregulation of neuronal nitric oxide synthase in the … of...key words: peripheral nerve injury,...

U

u

tK

K

M

Neuroscience 190 (2011) 367–378

UPREGULATION OF NEURONAL NITRIC OXIDE SYNTHASE IN THEPERIPHERY PROMOTES PAIN HYPERSENSITIVITY AFTER
PERIPHERAL NERVE INJURY
K. H. KIM,a J.-I. KIM,a,b J. A. HAN,c

M.-A. CHOEd* AND J.-H. AHNe*aDepartment of Biochemistry and Molecular Biology, Seoul National

niversity College of Medicine, Seoul 110-799, South KoreabDepartment of Biomedical Sciences, Seoul National University Grad-ate School, Seoul 110-799, South Korea

cDepartment of Biochemistry and Molecular Biology, Kangwon Na-ional University College of Medicine, Chuncheon 200-701, Southorea

dCollege of Nursing, Seoul National University, Seoul 110-799, Southorea

eDepartment of Biochemistry, Ewha Womans University School ofedicine, Seoul 158-710, South Korea

Abstract—Peripheral nerve injury often results in neuro-pathic pain that is manifested as hyperalgesia, and allodynia.Several studies suggest a functional role for neuronal nitricoxide synthase (nNOS) in the development or maintenance ofneuropathic pain, but such a contribution remains unclear. Inour current study, we found that intraplantar injection of theNOS substrate L-arginine or NO donor 3-morpholino-synoni-mine (SIN-1) produced mechanical hypersensitivity thatlasted more than 24 h. Following L5 spinal nerve ligation (L5SNL), immunoreactivity for nNOS in the ipsilateral L5 but notL4 dorsal root ganglion (DRG) was dramatically increased inmainly small- and medium-sized neurons and non-neuronalcells. L5 SNL caused increased nNOS immunoreactivity inthe ipsilateral sciatic nerve, mainly in Schwann cells and theipsilateral glabrous hind paw skin, mainly on the basementmembrane. Furthermore, total nNOS protein and mRNA in theipsilateral sciatic nerve and hind paw skin were markedlyupregulated following nerve injury. Intraplantar injection ofthe NOS inhibitor 7-nitroindazole (7-NI) or the non-specificNOS inhibitor L-NG-nitro-arginine methyl ester (L-NAME) ef-fectively suppressed SNL-induced mechanical allodynia. Col-lectively, these data suggest that in the periphery nNOS up-regulation induced by peripheral nerve injury contributes tomechanical hypersensitivity during the maintenance phaseof neuropathic pain. Blocking nNOS signaling in the periph-

*Corresponding author. Tel: �822-2650-5712 or �822-740-8824; fax:�822-2652-7848 or �822-765-4103.E-mail address: [email protected] (J.-H. Ahn) or [email protected](M.-A. Choe).Abbreviations: ATF3, activating transcription factor 3; CCI, chronicconstriction injury; cGMP, cyclic guanosine monophosphate; CGRP,calcitonin gene related peptide; D-NAME, D-NG-nitro-arginine methylester; DRG, dorsal root ganglion; GFAP, glial fibrillary acidic protein;IB4, isolectin 4; IL, interleukin; IR, immunoreactive; L-NAME, L-NG-nitro-arginine methyl ester; L5 SNL, L5 spinal nerve ligation; MPE,maximal possible effect; NF200, neurofilament 200; NMDA, N-methyl-D-aspartate, nNOS, neuronal nitric oxide synthase; NO, nitric oxide;NOS, nitric oxide synthase; PBS, phosphate buffer saline; PGP9.5,protein gene peptide 9.5; PWTs, paw withdrawal thresholds; SIN-1,
3-morpholino-synonimine; TRPV1, transient receptor potential cationchannel subfamily V member 1; 7-NI, 7-nitroindazole.
0306-4522/11 $ - see front matter © 2011 IBRO. Published by Elsevier Ltd. All righdoi:10.1016/j.neuroscience.2011.05.064

367

ery may thus be a novel therapeutic strategy for the treatmentof neuropathic pain. © 2011 IBRO. Published by Elsevier Ltd.All rights reserved.

Key words: peripheral nerve injury, mechanical allodynia,periphery, neuronal nitric oxide synthase, nitric oxide.

Peripheral nerve injury often results in hyperalgesia andallodynia, which are associated with neuropathic pain. Al-though many studies have made considerable progresstoward understanding neuropathic pain, the mechanismsunderlying neuropathic pain are poorly understood. In thepast, most reports have focused on only the neurons thatdrive the establishment and/or maintenance of neuropathicpain. However, recent investigations have demonstratedthe involvement of non-neuronal cells, such as immunecells and glial cells, in the pathogenesis of neuropathicpain (Oh et al., 2001; Ma and Eisenach, 2003; Zelenka etal., 2005; Thacker et al., 2009; Shibasaki et al., 2010).Nerve injury-induced events in non-neuronal cells canstimulate or recruit other cells or neurons, release a varietyof factors that are crucial for pain, and induce peripheralsensitization, thus directly increasing the excitability of no-ciceptors. These immune and glial cell responses to pe-ripheral nerve injury occur at several locations such asdorsal root ganglia (DRG), spinal cord, sciatic nerve, andperipheral sensory terminals (Takahashi et al., 2004;Zhuang et al., 2006; Scholz and Woolf, 2007; Shibasaki etal., 2010). However, much less is known about the inter-action between neuronal and non-neuronal cells in theperiphery under neuropathic pain conditions.

Nitric oxide (NO) is a diffusible molecule that acts as animportant modulator in the central and peripheral nervoussystems and that functions in various physiologic and patho-physiologic processes (Snyder, 1992; Meller and Gebhart,1993; Prast and Philippu, 2001). As NO activity is tightlyregulated by nitric oxide synthase (NOS), changes in expres-sion on NOS may regulate the pathophysiologic functions onNO in the nervous system. Neuronal NOS (nNOS) is ex-pressed in the neurons of the central and peripheral nervoussystem and predominantly produces NO in neuronal tissues.The contribution of nNOS to pain hypersensitivity has beencharacterized in neuropathic pain models (Luo and Cizkova,2000; Guan et al., 2007). First, nerve injury upregulatesnNOS expression in DRG neurons (Choi et al., 1996; Luo etal., 1999; Cizkova et al., 2002; Shortland et al., 2006; Guan etal., 2007), changes nNOS immunoreactivity in the spinal cord(Fiallos-Estrada et al., 1993; Steel et al., 1994; Cizkova et al.,
2002; Ma and Eisenach, 2007; Chacur et al., 2010), andts reserved.
mailto:[email protected]

mailto:[email protected]

n

K. H. Kim et al. / Neuroscience 190 (2011) 367–378368

alters the catalytic activity of nNOS in the DRG and spinalcord (Choi et al., 1996; Cizkova et al., 2002). In addition,genetic knockout of nNOS in mice attenuates pain hypersen-sitivity induced by nerve injury (Guan et al., 2007). It is sug-gested that NO synthesized by nNOS in the DRG or spinalcord following nerve injury activates protein kinase and/or ionchannels and thus results in neuronal excitability to causepain hypersensitivity (Qian et al., 1996; Kim et al., 2000;Yoshimura et al., 2001). However, pharmacological evidenceof specific and non-specific nNOS inhibitors administratedsystemically or spinally to alleviate nerve-injury induced painhypersensitivity remains controversial (Meller et al., 1992;Yoon et al., 1998; Luo et al., 1999; Pan et al., 1998; Lee et al.,2005; Guan et al., 2007; Chacur et al., 2010). This discrep-ancy can be explained by the differences in the deliverymethod, dose, drug potency or administration time afternerve injury. However, non-systemic or non-spinal adminis-tration can attenuate pain hypersensitivity induced by nerveinjury, but the analgesic effects of peripherally administratedspecific and non-specific nNOS inhibitor on neuropathic painhave not been reported.

In our current study, we hypothesized that upregulationof nNOS in the periphery of nerve-injured rats promotespain hypersensitivity. To test this hypothesis, we (1) askedwhether injection of an exogenous NOS substrate or NOdonor injection into the peripheral hind paw can inducemechanical hypersensitivity, (2) investigated nNOS ex-pression in DRGs, sciatic nerve, and hind paw skin afterperipheral nerve injury, and (3) examined the effects ofinjected nNOS or NOS inhibitors into the periphery onspinal nerve ligation (SNL)-induced mechanical allodynia.

EXPERIMENTAL PROCEDURES

These experiments conformed to the ethics guidelines of the Inter-national Association for the Study of Pain (IASP, 1983) and theNational Institutes of Health (USA). All procedures in this study wereaccorded with the National Institute of Health Guide for the Care andUse of Laboratory Animals (USA) and approved by the Seoul Na-tional University Hospital Animal Care and Use Committee. All effortswere made to minimize the number of animals used and theirsuffering.

Animals

Male Sprague–Dawley rats (230–250 g, Harlan) were used in allexperiments. The animals were housed in groups of two to fourper case, with food and water available ad libitum. All animalswere acclimated on 12-h light/dark cycle under standardized en-vironmental conditions.

Surgery

L5 SNL was performed as described previously (Jang et al.,2007). Briefly, all experimental procedures were performed underenflurane anesthesia (3% for induction and 2% for maintenance).A skin incision was made above the middle lumbar spine and theleft transverse process of L6 vertebra was identified. After care-fully removing the process, the L5 spinal nerve was isolated. Thenerve was tightly ligated with 6-0 silk thread and transected about1 mm distal to the ligation. The wound was aseptically sutured andmaintained with proper postoperative care. In sham operationgroup, these procedures were done in the same manner, except
for ligation and cut of the L5 spinal nerve.
Immunohistochemistry

Rats were anesthetized with sodium pentobarbital and transcar-dially perfused with 0.9% saline followed by 4% paraformaldehydein 0.1 M phosphate buffer saline (PBS), pH 7.4. The L4 and L5DRGs, sciatic nerve, and surface of hind paw skin were dissected.For sciatic nerve staining, about 50 mm of the sciatic nerve wastransected at the proximal portions of 50 mm apart from the regionwhere the distal sciatic nerve splits. The sciatic nerve was then cutlongitudinally or transversely. All tissues were post-fixed for 12 hat 4 °C in the same fixative and cryoprotected in 0.1 M PBScontaining 20% sucrose overnight at 4 °C. Samples weremounted in OCT and cryosectioned at 10 �m (DRGs and sciaticerve) or 30 �m (hind paw skin). For a single nNOS staining,

standard biotin-streptavidin techniques were used. Frozen sec-tions were washed three times with 0.3% Triton X-100 in 0.1 MPBS (T-PBS), incubated with 0.3% hydrogen peroxide for 15 min,and then blocked with 3% normal goat serum in T-PBS for 1 h.Samples were incubated with mouse anti-nNOS (1:1000; BD Bio-sciences, San Jose, CA, USA) overnight at 4 °C. After additionalwash, sections were incubated with biotinylated secondary anti-body (Zymed, San Francisco, CA, USA) for 1 h at room temper-ature and then applied with streptavidin-conjugated horseradishperoxidase (Zymed) for 15 min at room temperature. The specificnNOS binding was visualized with 3, 3-diaminobenzidine (DAB)(Zymed) and the sections were lightly counterstained with Hema-toxylin (Sigma, St. Louis, MO, USA).

For double immunofluorescent staining, frozen sections werewashed and blocked as mentioned above. Rabbit anti-calcitoningene related peptide (CGRP) (1:2000; Millipore, Billerica, MA,USA), transient receptor potential cation channel subfamily Vmember 1 (TRPV1) (1:1000; Millipore), activating transcriptionfactor 3 (ATF3) (1:400; Santa Cruz Biotechnology, Santa Cruz,CA, USA), glial fibrillary acidic protein (GFAP) (1:200; Sigma),S100 (1:400; Santa Cruz), or protein gene product 9.5 (PGP9.5)(1:1000; Abcam, Cambridge, UK) was incubated with mouse anti-nNOS (1:1000; BD) overnight at 4 °C. After an additional rise withT-PBS, the sections were incubated with anti-rabbit Alexa-Fluoro594 (1:500; Invitrogen, Carlsbad, CA, USA) and anti-mouseAlexa-Fluoro488 (1:500; Invitrogen) for 2 h at room temperature.For IB4 colocalization with nNOS, biotinylated isolectin 4 (IB4)(1:300; Sigma) and mouse anti-nNOS (1:1000; BD) were incu-bated overnight at 4 °C. After washing, the sections were incu-bated with anti-mouse Alexa-Fluoro488 (1:500; Invitrogen) andCY3-conjugated streptavidin (1:500; Sigma). For analysis of dou-ble-staining with neurofilament 200 (NF200) and nNOS, mousemonoclonal anti-NF200 antibody N52 (1:1000; Sigma) or mousemonoclonal anti-nNOS (1:1000; BD) were labeled with Alexa-Fluoro594 monoclonal antibody labeling kit (Invitrogen) or Alexa-Fluro488 monoclonal antibody labeling kit (Invitrogen), respec-tively. For glabrous hind paw and sciatic nerve samples, DAPIstaining was examined as control staining. The specificity of thestaining was confirmed by omitting of primary or secondary anti-bodies. Stained sections were visualized under LSM510 confocalmicroscope (Carl Zeiss Microscopy, Jena, Germany).

Quantitative analysis

Visualized DRG neurons with DAB were identified by the typicalmorphology in the presence of a nucleus. In each rat, four to sevensections of the L4/L5 DRG on day 7 post-SNL were randomly se-lected. Splitting of neuronal nuclei sections between sections canoverestimate true cell profiles. We, therefore, corrected split nucleiand calculated the neuronal number for each ganglion according toKonigsmark’s formula (Konigsmark, 1970). We divided the DRGneurons into small (�30 �m), medium (30–50 �m), and large (�50�m) neurons according to the mean of long- and short-axes of theneuronal soma (Harper and Lawson, 1985). Nuclear diameters were
measured as described previously (Jang et al., 2007). At least 300

ptDoew(pdm

91c

4HCrd

crPsus

n

st

snbmiAcaesabipsild

dp

K. H. Kim et al. / Neuroscience 190 (2011) 367–378 369

neuron profiles from each ganglion were measured and an assistantwho was unaware of the experimental group of the sections per-formed all counting procedures.

The subpopulations of double-labeling DRG neurons and sci-atic nerves were classified according to their labels and countedusing Adobe Photoshop CS (Adobe Systems Inc., San Jose, CA,USA). For quantifying double-labeling DRG neurons, five to sevensections of the ipsilateral L5 DRG on day 7 post-SNL were ran-domly selected and 1500–3000 cell profiles were counted in eachrat. Only neurons that did not overlap with other cells and withclearly visible nuclei were included in this counting. The averagepercentage of TRPV1-, IB4-, CGRP-, ATF3-, or NF200-IR neu-rons and nNOS-IR neurons, relative to the total number of neu-rons, was calculated for each animal across the different DRGsections and the counting data represented as mean�SEM. Forcounting nNOS-positive Schwann cells, three to four sections(100–200 �m apart) in each rat were randomly selected, theictures (112.5�112.5 �m (0.01265625 mm2)) were taken fromhese sections, and counting was performed from these pictures.omains of epineurium/perineurim, blood vessels, scar tissue andther non-neuronal tissues were excluded to analyze and only thendoneurial space was analyzed. S100-positive Schwann cellsere identified by shape and location among DAPI stained cells

Saito and Dahlin, 2008) and the average percentage of nNOS-ositive Schwann cells was calculated for each animal across theifferent sciatic nerve sections and the data represented asean�SEM.

Western blot

Total nNOS protein levels in the sciatic nerve and plantar surfaceof hind paw were individually compared at different time pointsafter nerve injury. All samples were homogenized with polytronhomogenizer (Kinematica) in lysis buffer containing 50 mM Tris–HCl (pH 8.0), 150 mM NaCl, 0.5% Triton X-100, 1 mM EDTA, 0.5mM DTT, 1 mM PMSF, 1 �g/ml pepstatin, 5 �g/ml leupeptin, and

�g/ml aprotinin. Insoluble debris was cleared by centrifusion at5,000�g for 30 min at 4 °C and then the supernatant wereollected as extracts. Equal amounts of protein (30 �g/per lane)

were separated from 10% SDS-PAGE and transferred to nitrocel-lulose membrane. The blots were blocked with 0.1% Tween 20containing 5% milk in PBS and incubated with mouse anti-nNOS(1:1000; BD) or mouse anti-� actin (1:5000; Sigma) overnight at

°C. After washing, the specific binding was detected usingRP-conjugated anti-mouse secondary antibodies (1:3000; Santaruz) for 1 h at room temperature followed by chemiluminescent

eagents. The quantification of specific bands was performed byensitometric analysis using MultiGauge software (Fujifilm).

Real-time RT-PCR

In order to investigate temporal changes of nNOS transcripts afternerve injury, sciatic nerve and plantar surface of hind paw wereindividually obtained at different time points after nerve injury. Thetissues were homogenized by polytron homogenizer and extractedwith Trizol (Invitrogen). Total RNA was isolated by manufacturer’ssuggesting protocol and rapidly frozen with dry ice and stored at�80 °C until ready to use. Total RNA was reverse transcribed intoDNA using SuperScript III reverse transcriptase (Invitrogen) andandom hexamer primer. PCR primers for nNOS (target) and RNase

(single copy number gene as an endogenous control) were de-igned for coding region of genes and its specificity was assessedsing blast search at NCBI. The primers were as follows: nNOSense 5=-CTTTTCATCGTGGGGTCAAT-3= and antisense 5=-AAG-

CCGTCGATCTGTCTCAC-3=; RNase P sense 5=-TTCACTGC-TTCATGCCTACG-3= and antisense 5=-GTTGGTTCAGTCCGAT-GCTT-3=. Subsequent PCR was carried out in 250 nmol primers, 500ng cDNA, and 2� SYBR Premix Ex Taq II (Takara) with the IQ5
Multicolor Real-time PCR Detection System (Bio-Rad). Amplification
conditions were 50 °C for 2 min, 95 °C for 10 min and 40 cycles at95 °C for 15 s, and 60 °C for 45 s followed by dissociation curveanalysis to confirm the specificity. After completing the PCR, theamount of nNOS mRNA in duplicates was estimated by 2-��Ctformula as follows: ��Ct�[Ct target (unknown sample)�Ct endog-enous control (unknown sample)]�[Ct target (calibrator sample)�Ctendogenous control (calibrator sample)]. Calibrator samples wereobtained from sham-operated rats.

Drug administration

3-morpholino-synonimine (SIN-1), L-arginine, D-arginine, L-NG-argi-ine methyl ester (L-NAME), and D-NG-arginine methyl ester (D-

NAME) were obtained from Sigma and 7-Nitroindazole (7-NI) waspurchased from Cayman Chemicals. 7-NI was dissolved in dimeth-ylsulfoxide (DMSO) and SIN-1, L-arginine, D-arginine, L-NAME, orD-NAME was prepared in PBS. 7-NI, L-NAME, D-NAME, SIN-1,L-arginine, or D-arginine was initially dissolved at the concentration of500 �M, 100 �M, 100 �M, 10 mg/ml, 50 mg/ml, or 50 mg/ml as stockolution, respectively and then diluted with PBS to the final concen-ration.

Pro-nociceptive effect of nitric oxide was investigated with SIN-1,L-arginine, or D-arginine injected into perisciatic nerve or plantarsurface of the hind paw in a volume of 20 �l using 50 �l Hamiltonyringe with 30 gauge needle. For a single injection into the sciaticerve, left sciatic nerve at mid-thigh level was aseptically exposed bylunt dissection under enflurane anesthesia. After grasping adjacentuscle with forceps to secure a view, the solution was slowly injected

nto the sciatic nerve and the needle was then carefully removed.fter confirming no backflow, the wound was carefully and asepti-ally sutured. For the injection into the sciatic nerve, left sciatic nervet mid-thigh level was aseptically exposed by blunt dissection undernflurane anesthesia. After grasping adjacent muscle with forceps toecure a view, the solution was slowly injected into the sciatic nervend the needle was then carefully removed. After confirming noackflow, the wound was carefully and aseptically sutured. For the

ntraplantar injection, the needle was inserted into the left plantar skinroximal to the midpoint of the hind paw and advanced about 10 mmo that it reached the midpoint of the paw. The solution was slowlynjected forming a bleb that disappeared within the 10 min and anyeakage was not observed. All procedures were performed under aissection microscope.

Analgesic effect of neuronal nitric oxide on nerve injury-in-uced mechanical allodynia was examined peri-sciatic or intra-lantar injection. Peri-sciatic or intraplantar injection of 7-NI, L-

NAME or PBS was performed on day 3 post-SNL as describedabove. No drug induced stereotypical alterations in motor activityincluding hyper-locomotion or balance loss.

Behavioral tests

Mechanical hypersensitivity was tested by measuring paw with-drawal thresholds (PWTs) with the application of a von Frey filamentusing the up-down paradigm. Rats were placed in plexiglass caseabove a wire mesh bottom that allowed full access to their paws.After a habituation period, the series of von Frey filaments from 0.4 to15.0 g (0.4, 0.6, 1.0, 1.4, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0 g) (North coast)were applied perpendicular to the middle of the plantar surface of thehind paw for 2–3 s. The 2.0 g stimulus was applied first. The filamentwas applied with at least 3-min interval between stimuli and calcu-lated as previously described (Chaplan et al., 1994). All behavioraltests were carried out by experimenters blinded to drug treatmentsand preceding surgeries.

Statistical analysis

All data were given as means�SEM. Variances between ipsilat-eral and contralateral side on the same group were compared by
Wilcoxon signed-rank test. Differences between more than three

�o0r

ss

s

1rh2

ut testinat 2 h af


different groups were evaluated by Kruskall-Wallis test. Statisticalanalysis on the variances between two different groups was per-formed by Mann–Whitney rank-sum test for unmatched pairs. Thedata were considered significant if P-value is less than 0.05.

For investigating the dose-response functions and comparingdrug effects on SNL-induced mechanical allodynia, MPE (maximumpossible effect) value was analyzed: MPE (%)�[1�(cut-off PWT-

post-drug PWT)/(cut-off PWT�pre-drug PWT)]�100, where cut-ff PWT is 18 (g). The resulting MPE (%) are continuous data from% to 100% and ED50 (dose estimated to produce 50% MPE) foreversing mechanical allodynia was calculated accordingly.

RESULTS

Pro-nociceptive effects of intraplantar NOS substrateand NO donor injection

Paw withdrawal thresholds were investigated following intra-

Fig. 1. Mechanical hypersensitivity produced by intraplantar injection5 min after intraplantar injection of NOS substrate L-arginine (0.5 �g;espectively (� P�0.05 vs. saline injection group; n�7). Intraplantar injypersensitivity in the ipsilateral side of paw. (B) Mechanical threshold0 �g L-arginine or 0.2, 2.0, and 20 �g SIN-1 (� P�0.05 vs. saline inj

nor SIN-1 (5.0 �g; n�6) induces mechanical hypersensitivity throughoinjected into the sciatic nerve cause no altered mechanical sensitivity

plantar injection (Fig. 1A, B) or perisciatic injection (Fig. 1C, (

D) of the NOS substrate L-arginine or NO donor SIN-1. Intra-plantar injection of L-arginine (0.5 �g) or SIN-1 (5.0 �g)significantly reduced mechanical withdrawal thresholds com-pared with saline injection (Fig. 1A). The mechanical hyper-sensitivity began at 15 min after injection and continued formore than 24 h in ipsilateral paws. Intraplantar injection ofneither saline nor D-arginine induced mechanical hypersen-itivity. Withdrawal thresholds on the contralateral sidehowed no significant changes after SIN-1 or L-arginine in-

jection (data not shown). Next, the dose-dependent hyperal-gesic effects of SIN-1 or L-arginine were investigated. Ashown in Fig. 1B, intraplantar injection of L-arginine (0.2–20

�g) or SIN-1 (0.2–20 �g) caused significant mechanical hy-persensitivity at 2 h after administration. In contrast, perisci-atic nerve injection of neither L-arginine (0.5 �g) nor SIN-1

OS substrate or NO donor. (A) Withdrawal thresholds are reduced atthe NO donor SIN-1 (5.0 �g; n�7) and these last up to 24 h or 48 h,neither D-arginine (0.5 �g; n�6) nor saline (n�7) induces mechanicalnificantly decreased at 2 h after intraplantar injection of 0.2, 2.0, andup; n�6). (C) Perisciatic injection of neither L-arginine (0.5 �g; n�8)

g period compared with saline-injected group. (D) L-arginine or SIN-1ter administration.

of the Nn�7) or

ection ofs are sig

ection gro

5.0 �g) produced significant pain hypersensitivity during

Sc

InD

nDSaFscr

ui

pared witlateral sid


testing period (Fig. 1C). Perisciatic injection of L-arginine orIN-1 at high dosage did not alter the mechanical sensitivityompared with vehicle (Fig. 1D).

ncreased expression of nNOS in primaryociceptive neurons and nonneuronal cells of L5RG following L5 SNL

NOS immunoreactivity was increased in the ipsilateral L5RG compared with the contralateral side at 7 days afterNL (Fig. 2A, B). nNOS immunoreactivity was investigatedccording to cell size based on cell diameter. As shown inig. 2C, nNOS immunoreactivity was mainly detected inmall- and medium-sized neurons in the ipsilateral L5 DRGompared with the contralateral side. Among 504�33 neu-ons in the ipsilateral L5 DRG of nerve-injured rats (n�3),

138�17 were nNOS-positive neurons; of these, 39% weresmall, 19% were medium, and 3% were large-sized neurons.In contrast, SNL induced no increase in nNOS immunoreac-tivity in the ipsilateral L4 DRG compared with the contralateralside (Fig. 2D, E). Among 513�19 neurons in the ipsilateral L4DRG of nerve-injured rats (n�3), 43�7 were nNOS-positiveneurons; of these, 12% were small, 9% were medium, and1% were large-sized neurons (Fig. 2F).

Next, the subpopulations of ipsilateral L5 DRG neuronsdisplaying nNOS immunoreactivity were examined at 7days after surgery. Fig. 3A shows typical examples ofnNOS-positive neurons coexpressed with TRPV1 (amarker of primary nociceptive neurons), IB4 (a marker ofnonpeptidergic nociceptors), CGRP (a marker of peptider-

Fig. 2. Increased nNOS immunoreactivity in injured L5 but not unipregulated in ipsilateral L5 DRG neurons compared with the contrala

ncreased in small- and medium-sized neurons of the ipsilateral L5 DnNOS expression is not increased in ipsilateral L4 DRG neurons comany cell type in ipsilateral L4 DRG neurons compared with the contra

gic nociceptors), ATF3 (a marker of damaged primary

neurons), NF200 (a marker of myelinated primary neu-rons), or GFAP (a marker of satellite glial cells). nNOSimmunoreactivity in L5 DRG neurons colocalized with thenociceptive neuron markers, TRPV1, IB4, and CGRP, andthe myelinated neuron marker, NF200. Moreover, nNOSand ATF3 were strongly colocalized. Interestingly, nNOSimmunoreactivity was clearly observed in non-neuronalcells such as satellite glial cells (insets in middle column)following L5 SNL. Most nNOS positive satellite cells werenon-neuronal cells surrounding neuronal cells that werenNOS negative. Double staining for nNOS and GFAPshowed that the majority of nNOS-positive non-neuronalcells in the L5 DRG colocalized with GFAP.

For all double-labeling experiments, the ipsilateral L5DRGs from three rats (five to seven sections per rat) onday 7 post-SNL were analyzed. The percentages of nNOS-immunoreactive (IR) neurons labeled with various markerswere as follows: TRPV1, 33�4; IB4, 28�4; CGRP, 18�3;ATF3, 62�4; NF200, 10�1 (Fig. 3B).

SNL-induced nNOS upregulation in the sciatic nerveand plantar surface of the hind paw

Next, nNOS expression following SNL in the periphery wasinvestigated with double-staining. As shown in Fig. 4A,strong nNOS immunoreactivity was observed in longitudi-nally sectioned ipsilateral sciatic nerve compared with thecontralateral side of nerve-injured rats. Nerve injury-in-duced nNOS upregulation rarely colocalized with NF200 inthe ipsilateral sciatic nerve compared with the contralateral

4 DRG following L5 SNL. (A, B) nNOS immunoreactivity is clearlye at 7 d after nerve injury. (C) nNOS immunoreactivity is significantlywing peripheral nerve injury (� P�0.05 vs. contralateral side). (D, E)h the contralateral side. (F) nNOS immunoreactivity is not changed ine. Scale bar, 50 �m.

njured Lteral sidRG follo

side. Interestingly, the weak but widely distributed nNOS

inmScn

flmaeHoaaf

ttpdi

ns co-labOS-IR is


immunoreactivity was observed in dense cellular clustersof ipsilateral sciatic nerve following nerve injury. Moreover,nNOS in the transversely sectioned ipsilateral sciatic nervestrongly colocalized with S100, a marker of Schwann cells,at 7 days after nerve injury (Fig. 4B). nNOS also colocal-zed with ATF3 in longitudinally sectioned ipsilateral sciaticerve. However, unlike ATF3, nNOS immunoreactivity wasostly observed in Schwann cells but not in axons afterNL (Fig. 4B). The percentages of nNOS-immunoreactiveells colocalized with S100 was 71�9 in ipsilateral sciaticerves and 7�2 in contralateral sciatic nerves (n�3).

Glabrous hind paw skin was examined with immuno-uorescent staining with PGP9.5, a panneuronal fiberarker, and nNOS. As shown in Fig. 5, nNOS immunore-ctivity on the contralateral side was rarely found in thepidermis or dermis of hind paw skin on day 7 post-SNL.owever, nNOS was clearly induced on the ipsilateral sidef glabrous hind paw skin. In particular, nNOS immunore-ctivity was mostly found on the basement membrane anddjacent PGP9.5-positive nerve fibers in the plantar sur-ace of the hind paw.

nNOS protein (Fig. 6A) and mRNA (Fig. 6B) levels overime were evaluated with western blot analysis and real-ime RT-PCR, respectively. As shown in Fig. 6A, nNOSrotein in the ipsilateral sciatic nerve was increased at 3ays, and gradually increased up to 14 days after nerve

Fig. 3. Increased nNOS expression in primary nociceptive neuronsConfocal images show typical examples of nNOS immunoreactivity iTRPV1, IB4, CGRP, ATF3, NF200 or GFAP. Arrows indicate doubleimages of the areas enclosed by the dashed boxes show pronounced nof middle columns). Moreover, nNOS expression of the ipsilateral L5(C) For all bars, green bar represents the nNOS-IR and red bar showthe percentage of labeled neurons. The percentage of nNOS-IR neuropercentage of specific marker-positive neuron double-labeled with nN

njury. In the ipsilateral hind paw skin, nNOS expression

was upregulated at 3 days, peaked at 7 days, and wasslightly decreased at 14 days after SNL. The nNOS proteinlevels in the sciatic nerve and hind paw skin of SNL-injuredrats (n�4, each) were individually quantified by measuringthe density of the nNOS bands and are shown as the meanpercent of levels in sham-operated rats (n�4, each). Sci-atic nerve and hind paw skin of sham-operated rats wereobtained at 7 days after surgery.

The levels of nNOS transcripts from the ipsilateralsciatic nerve and hind paw skin of rats (n�4, per timepoint) were individually normalized to levels of a single-copy number gene (RNase P) and evaluated with the ��Ctmethod (Fig. 6B). In the ipsilateral sciatic nerve, nNOSmRNA was increased at 3 days, peaked at 7 days, andwas slightly decreased at 14 days following nerve injury. Inthe ipsilateral hind paw skin, nNOS mRNA was upregu-lated at 3 days after L5 SNL.

Effects of nNOS inhibitor treatment in the peripheryon SNL-induced mechanical allodynia in theperiphery

L5 SNL rapidly decreased PWTs, which were maintainedat decreased levels. The PWTs of the affected hind pawwere decreased compared with the contralateral hind pawand sham-operated rats, and lasted throughout the testing

-neuronal cells of the ipsilateral L5 DRG following nerve injury. (A)ulations of the ipsilateral L5 DRG double-labeled with antibodies topsilateral L5 DRG neurons on day 7 post-SNL. Higher-magnificationunoreactivity in non-neuronal cells of the L5 DRG (arrowheads; insetsvily colocalized with GFAP compared with the contralateral side (B).PV1-, IB4-, ATF3-, or NF200-IR. The number in each bar representseled with marker is shown in the yellow area of top bar, whereas theshown in the yellow area of bottom bar. Scale bar, 20 �m.

and nonn subpop-labeled iNOS immDRG heas the TR

period (Fig. 7A). To determine whether nNOS in the sciatic

dpt

m.


nerve or peripheral hind paw is involved in the mainte-nance of neuropathic pain, we injected 7-NI or L-NAME intothe sciatic nerve or hind paw skin on day 3 post-SNL

Fig. 4. nNOS upregulation in the sciatic nerve at 7 d after peripheralnNOS on the ipsilateral side compared with the contralateral side. Connerve following nerve injury. Arrow indicates clearly nNOS-positivenNOS-immunoreactivity in the dense cellular clusters. Scale bar, 2immunoreactivity using S100 (Schwann cell marker) or ATF3 (injured nsciatic nerve following nerve injury compared with contralateral side. Hin sciatic nerve. Arrows indicate double-labeled Schwann cells in theand ATF3-positive cells in the ipsilateral sciatic nerve. Scale bar, 20 �

(n�9–11). Peri-sciatic administration of 7-NI or L-NAME m

id not reverse SNL-induced mechanical allodynia com-ared with saline treatment (Fig. 7B). However, intraplan-ar administration of 7-NI and L-NAME effectively reversed

jury. (A) Longitudinally sectioned sciatic nerve shows upregulation ofges demonstrate rare colocalization of nNOS with NF200 in the sciatiche ipsilateral sciatic nerve. Arrowhead shows weak but distributed

) Double-labeling images demonstrate typical examples of nNOSker) antibodies. nNOS strongly colocalizes with S100 in the ipsilateralnNOS expression is found on both ATF3-positive and –negative cellsl sciatic nerve at 7 d after L5 SNL. Arrowheads show nNOS-negative

nerve infocal imacell in t0 �m. (Berve marowever,

ipsilatera

echanical pain hypersensitivity for �3 h after injection

s indicate


compared with saline treatment (Fig. 7C). The dose–re-sponse effect of 7-NI and L-NAME is shown in Fig. 7D.MPE was calculated at 2 h after intraplantar injection onday 3 post-SNL, and then ED50 was calculated. The ED50

values were 73.4 and 486.0 nmol, respectively (Fig. 7D).Intraplantar injection of 7-NI or L-NAME into the contralat-eral hind paw even at high dose did not reverse SNL-induced mechanical allodynia (data not shown).

DISCUSSION

Peripheral nerve injury has distinct effects on nNOS ex-pression in the DRG and spinal cord. In chronic constric-tion injury (CCI) model rats, nNOS immunoreactivity isincreased in the ipsilateral L4-L6 DRG (Cizkova et al.,2002). In L5 SNL animal models, nNOS protein is upregu-lated in the ipsilateral L5 DRG (Shortland et al., 2006;Guan et al., 2007) and nNOS protein and mRNA areincreased in the ipsilateral L5-L6 DRG (Luo et al., 1999).Increased nNOS expression is mainly detected on small-and occasionally medium-sized neurons in the DRG (Luoet al., 1999; Shortland et al., 2006). Consistent with previ-ous studies, we detected marked nNOS expression inipsilateral neurons, mainly in small- and medium-sizedneurons, of the L5 DRG after L5 SNL. Moreover, ourconfocal images demonstrated that nNOS-positive neu-rons also expressed primary afferent nociceptors followingnerve injury. In particular, nNOS in the injured L5 DRGstrongly colocalized with ATF3. ATF3, which is widelyused as a marker of injured neurons, forms a complexwith another transcription factor, c-jun, and regulates orchanges protein expression, modulating neuronal sensitiv-ity (Tsujino et al., 2000; Ji and Strichartz, 2004). Interest-

Fig. 5. nNOS induction in the plantar surface of glabrous hind paw fopanneuronal fiber marker, in the contralateral side and the ipsilateral sin the epidermis or dermis of the contralateral glabrous skin. However,is mostly found on the basement membrane, an epidermis-dermis borddistributions throughout the epidermis and dermis of paw skin. Arrow

ingly, we observed nNOS upregulation in non-neuronal

cells as well as neuronal cells in the ipsilateral L5 DRGfollowing nerve injury compared with the contralateral side.Our report is the first to demonstrate nNOS upregulation innon-neuronal cells such as satellite cells in the injured L5DRG after L5 SNL. Recent investigations have shown theinvolvement of immune function on DRGs in the establish-ment and/or maintenance of neuropathic pain. In particu-lar, peripheral nerve injury induces increased communica-tion between satellite cells and neurons and thus amplifiesneuronal excitability and enhances primary afferent inputs(Scholz and Woolf, 2007; Ren and Dubner, 2010). More-over, activated chemokines and cytokines regulate theexpression of other proteins on non-neuronal cells andinduce secretion of other proteins from injured neurons.For example, matrix metalloproteinase 2 (MMP2) is up-regulated in satellite cells via interleukin (IL)-1 activation,and an MMP2 inhibitor effectively attenuates SNL-inducedmechanical hypersensitivity (Kawasaki et al., 2008).

Although nNOS expression and maintenance and/ordevelopment of neuropathic pain are tightly correlated, noother reports have shown nNOS expression in sciaticnerve following L5 SNL. We found marked nNOS upregu-lation in the ipsilateral sciatic nerve after SNL and thisexpression rarely detected in NF200-positive neurons.These results suggest that upregulated nNOS in the sciaticnerve following nerve injury may not be transported fromthe DRG via axons into the periphery. Instead, nNOSexpression was strongly detected in non-neuronal cells(mainly Schwann cells) in the ipsilateral sciatic nerve afternerve injury. Unlike nNOS, ATF3 immunoreactivity wasfound on both neuronal cells in axons and Schwann cellsin the ipsilateral sciatic nerve. Thus, the origin of the up-

NL. Confocal images show immunoreactivity of nNOS and PGP9.5, aabrous hind paws at 7 d following L5 SNL. nNOS is rarely expressedthe ipsilateral side is increasingly expressed and the immunoreactivityhead). Nuclei are highlighted with pseudocolored gray to show cellulartypical example of PGP9.5-positive nerve fiber. Scale bar, 50 �m.

llowing Side of gl

nNOS iner (arrow

regulated nNOS may be activated Schwann cells in the

aSgssnoccsc

nNis

SpScpjeifoiaidiolbmh

ps

hind paw


sciatic nerve following nerve injury. Moreover, total levelsof nNOS protein and transcripts were increased in theipsilateral sciatic nerve following nerve injury (Fig. 6A, B).During Wallerian degeneration following nerve injury,Schwann cells enveloping degenerating axons undergomarked reactive changes, begin to phagocytose myelindebris, and synthesize several biological molecules, in-cluding nerve growth factor, tumor necrosis factor-�, IL-1�,nd IL-6 (Matsuoka et al., 1991; Murwani et al., 1996;hamash et al., 2002). In this study, we could not investi-ate the origin or mechanism of the nNOS induction in theciatic nerve, but it may result from the activation of apecific receptor in Schwann cells leading to modulation ofNOS expression via a transcription pathway such as thatf c-Jun N-terminal kinase (JNK) or p38 kinase, or in-reased synthesis of a biological molecule in Schwannells following nerve injury that modulates nNOS expres-ion via interaction with proinflammatory cytokines orhemokines (Oh et al., 2001; Scholz and Woolf, 2007).

Unlike upregulation of nNOS in sciatic nerve followingerve injury, peri-sciatic nerve injection of an NO donor orOS substrate evoked no mechanical pain hypersensitiv-

ty (Fig. 1C, D). Moreover, peri-sciatic nerve injection of a

Fig. 6. Increased nNOS protein and transcripts in the sciatic nerve aipsilateral sciatic nerve is increased at 3 d and gradually upregulatedin the ipsilateral hind paw skin is increased at 3 d, peaks at 7 d and is smRNA in the ipsilateral sciatic nerve is increased 3 d and remains up tvs. sham-operated rats). The nNOS mRNA level in the ipsilateralsham-operated rats).

pecific or non-specific nNOS inhibitor did not alleviate (

NL-induced pain hypersensitivity (Fig. 7B). Repeatederi-sciatic nerve injection, but not a single injection, inNL-induced rats can alleviate pain hypersensitivity, buthronic peri-sciatic administration with a catheter causes aeri-sciatic immune reaction and repeated peri-sciatic in-

ection itself induces damage on the sciatic nerve (Milligant al., 2003). SNL causes Wallerian degeneration in the

psilateral sciatic nerve, where multiple factors releasedrom the injured nerve fibers of L5 DRGs affect the functionf intact adjacent nerve fibers of L4 DRGs. nNOS induction

n the sciatic nerve (mainly in Schwann cells) may linkctivations of several cellular functions, including activated

on channels. Thus only blocking nNOS in the sciatic nerveuring the maintenance phase may be ineffective in reduc-

ng neuropathic pain. Furthermore, a peri-sciatic injectionf NO donor or NOS substrate can affect the sciatic nerve

ocally or in a limited way and the modulated NO signalingy peri-sciatic injection in the sciatic nerve cannot be trans-itted or linked to peripheral target tissues, inducing painypersensitivity.

Conflicting pharmacological evidence has been re-orted concerning the analgesic effect of specific or non-pecific nNOS inhibitors on neuropathic pain. Intrathecal

paw skin of nerve-injured rats. (A) Total level of nNOS protein in thed after nerve injury (� P�0.05 vs. sham-operated rats). nNOS proteincreased 14 d after SNL (� P�0.05 vs. sham-operated rats). (B) nNOSeaking at 7 d and decreasing at 14 d following nerve injury (� P�0.05

skin is significantly upregulated at 3 d after SNL (� P�0.05 vs.

nd hindup to 14lightly deo 14 d, p

Meller et al., 1992; Guan et al., 2007; Chacur et al., 2010)

(pA

Lo tested a


or intraperitoneal (Yoon et al., 1998; Guan et al., 2007)administration of specific and/or non-specific nNOS inhib-itors effectively reversed neuropathic pain, but other inves-tigators reported that intrathecal or intraperitoneal admin-istration of specific and/or non-specific nNOS inhibitorshas no analgesic effects on neuropathic pain (Luo et al.,1999; Pan et al., 1998; Lee et al., 2005). Khalil and Khodr(2001) reported that intramuscular injection of the specificnNOS inhibitor 7-NI into the mid-thigh region of the sciaticnerve effectively reduces CCI-induced thermal hyperalge-sia. Direct injection of non-specific nNOS inhibitor L-NAMEinto the injured sciatic nerve also alleviates CCI-inducedhyperalgesia, but systemic L-NAME injection does not re-duce the pain hypersensitivity following CCI (Thomas etal., 1996). As shown in Fig. 7, our study first reports ananalgesic effect of intraplantar injection of 7-NI or L-NAMEon SNL-induced pain hypersensitivity. Moreover, total

Fig. 7. Analgesic effect of nNOS or NOS inhibitor injected into plantardecrease in paw withdrawal thresholds (PWTs) on the ipsilateral sideReduction in the mechanical stimuli is observed 1 d after L5 SNLnNOS-specific inhibitor 7-NI or non-specific NOS inhibitor L-NAME evsaline treatment (n�9, each). (C) Intraplantar injection of 7-NI (100 nm5 SNL (� P�0.05 vs. saline injected rats (n�10–11). (D) 7-NI and L-NAn the SNL-induced mechanical allodynia. Mechanical thresholds were

486.0 nmol, respectively.

nNOS protein, transcripts, and immunoreactivity were in- i

creased in the ipsilateral hind paw skin after nerve injury(Figs. 5 and 6). In particular, nNOS immunoreactivity in thehind paw skin was mainly detected on basement mem-brane, near primary afferent terminals. Considerable evi-dence of clinical studies has been shown the role of NO innociception under pain status (Christiansen et al., 1999;Kimura et al., 1999; Takahashi et al., 1999). In particular,intracutaneous injection of NO precursor evokes pain inhumans (Holthusen and Arndt, 1995), and intradermal in-jection of an NO donor or NOS substrate produces me-chanical hyperalgesia in rats (Aley et al., 1998). Consistentwith these reports, we observed mechanical hypersensi-tivity following intraplantar administration of the NOS sub-strate L-arginine (0.5 �g) or the NO donor SIN-1 (5.0 �g)Fig. 1). The hypersensitivity began at 15 min after intra-lantar injection and continued for up to 2 days. However,ley et al. (1998) reported that mechanical hyperalgesia

on SNL-induced mechanical allodynia. (A) L5 SNL leads to significantP�0.05 vs. contralateral side and sham-operated rats (n�7, each).

ts throughout the testing period. (B) Peri-sciatic administration ofnalgesic effect on SNL-induced mechanical allodynia compared withAME (500 nmol) effectively reversed mechanical allodynia at 3 d afterplantarly injected on day 3 post-SNL, showed dose-dependent effectst 2 h after injection. The ED50 value for 7-NI and L-NAME is 73.4, and

surfaceof rats (�and las

okes no aol) or L-NME, intra

nduced by intradermal administration of the NO donor


SIN-1 (10 �g) or the NOS substrate L-arginine (40 �g) intothe paw lasts less than 4 h. The discrepancy may be dueto the delivery method, drug potency, dosage, or adminis-tration times. For example, we injected SIN-1 or L-arginineinto the plantar skin proximal to the midpoint of the hindpaw and advanced the needle about 10 mm so that itreached the midpoint of the paw. Because Aley et al.(1998) did not clearly state the injection method, we cannotexplain this discrepancy with this study.

Recently, the role of nNOS on neuropathic pain wasdemonstrated using nNOS knockout mice. nNOS knockoutmice shows blunted SNL-induced mechanical hypersensi-tivity, and intrathecal injection of a specific nNOS inhibitoralleviates mechanical hypersensitivity after nerve injury inwild type mice (Guan et al., 2007). In addition, total nNOSprotein in L5 DRG (but not spinal cord) was upregulatedafter L5 SNL, and thus, Guan et al. (2007) suggested thatnNOS in the DRG participates in the development and/ormaintenance of mechanical hypersensitivity followingnerve injury. Although the role of nNOS in the spinal cordor DRG on the development and maintenance of neuro-pathic pain has been demonstrated, the contribution ofnNOS in the periphery to neuropathic pain is unclear. It iswell known that in the central nervous system, nNOSsynthesized-NO following the influx of Ca2� through N-methyl-D-aspartate (NMDA) receptors implicates synapticplasticity in the spinal cord (Meller and Gebhart, 1993).Activated NMDA receptors coupled with nNOS lead tofurther release of excitatory neurotransmitters, resulting inpositive feedback regulation of neuronal hyperexcitabilityin the dorsal horn after nerve injury (Luo and Cizkova,2000; Ji et al., 2003). In the periphery, intraplantar injectionof non-competitive NMDA receptor antagonist MK-801also delays the onset of neuropathic pain and attenuatesnerve injury-induced mechanical pain hypersensitivity(Jang et al., 2004). Jang et al. (2004) used a modifiedsurgical method, L5 SNL preceded by L5 dorsal rhizotomy,to avoid the potential central effects in L5 SNL animalmodels. Moreover, some reports showed an analgesiceffect of specific and non-specific nNOS inhibitors on neu-ropathic pain with intrathecal administration (Meller et al.,1992; Guan et al., 2007; Chacur et al., 2010). As the cellsin the spinal cord are enveloped by the subarachnoidspace, intrathecally injected drugs can act peripherally aswell as spinally (Gendron et al., 2006; Chacur et al., 2010).However, the contribution of nNOS in the periphery toneuropathic pain has been overlooked. Generally, NO as agaseous molecule diffuses out from the DRG neurons orprimary afferent terminals, binds to soluble guanylyl cy-clase, and generates cyclic guanosine monophosphate(cGMP). cGMP modulates various neuronal functions byactivating protein kinases, ion channels, or inflammatorymediators (Meller and Gebhart, 1993; Prast and Philippu,2001). Therefore, it is likely that nerve injury induces up-regulation of peripheral nNOS which is linked with acti-vated NMDA receptors in the periphery, and subsequentlyreleased NO acts as second messenger to facilitate neu-
ronal excitability.
In conclusion, we show that intraplantar injection of anNO donor or NOS substrate provoked lasting mechanicalhypersensitivity. nNOS was clearly upregulated in neuro-nal and non-neuronal cells of the DRG and sciatic nerve,and hind paw skin following peripheral nerve injury. Treat-ment with inhibitors of nNOS or NOS into the peripheralhind paw of rats effectively attenuated SNL-induced me-chanical allodynia. nNOS is thus a pivotal factor for themaintenance of pain hypersensitivity following nerve injury,and blocking nNOS signaling may be a promising thera-peutic strategy for treatment of neuropathic pain.

Acknowledgments—This work was supported by a Korea Scienceand Engineering Foundation (KOSEF) grant AQ2 funded by theKorean government (MOST; R01-2007-000-10573-0).

REFERENCES

Aley KO, McCarter G, Levine JD (1998) Nitric oxide signaling in painand nociceptor sensitization in the rat. J Neurosci 18:7008–7014.

Chacur M, Matos RJ, Alves AS, Rodrigues AC, Gutierrez V, Cury Y,Britto LR (2010) Participation of neuronal nitric oxide synthase inexperimental neuropathic pain induced by sciatic nerve transec-tion. Braz J Med Biol Res 43:367–376.

Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL (1994)Quantitative assessment of tactile allodynia in the rat paw. J Neu-rosci Methods 53:55–63.

Choi Y, Raja SN, Moore LC, Tobin JR (1996) Neuropathic pain in ratsis associated with altered nitric oxide synthase activity in neuraltissue. J Neurol Sci 138:14–20.

Christiansen I, Thomsen LL, Daugaard D, Ulrich V, Olesen J (1999)Glyceryl trinitrate induces attacks of migraine without aura in suf-ferers of migraine with aura. Cephalalgia 19:660–667; discussion626.

Cizkova D, Lukacova N, Marsala M, Marsala J (2002) Neuropathicpain is associated with alterations of nitric oxide synthase immu-noreactivity and catalytic activity in dorsal root ganglia and spinaldorsal horn. Brain Res Bull 58:161–171.

Fiallos-Estrada CE, Kummer W, Mayer B, Bravo R, Zimmermann M,Herdegen T (1993) Long-lasting increase of nitric oxide synthaseimmunoreactivity, NADPH-diaphorase reaction and c-JUN co-ex-pression in rat dorsal root ganglion neurons following sciatic nervetransection. Neurosci Lett 150:169–173.

Gendron L, Lucido AL, Mennicken F, O’Donnell D, Vincent JP, Stroh T,Beaudet A (2006) Morphine and pain-related stimuli enhance cellsurface availability of somatic delta-opioid receptors in rat dorsalroot ganglia. J Neurosci 26:953–962.

Guan Y, Yaster M, Raja SN, Tao YX (2007) Genetic knockout andpharmacologic inhibition of neuronal nitric oxide synthase attenu-ate nerve injury-induced mechanical hypersensitivity in mice. MolPain 3:29.

Harper AA, Lawson SN (1985) Conduction velocity is related to mor-phological cell type in rat dorsal root ganglion neurones. J Physiol359:31–46.

Holthusen H, Arndt JO (1995) Nitric oxide evokes pain at nociceptorsof the paravascular tissue and veins in humans. J Physiol 487 (Pt1):253–258.

Jang JH, Kim DW, Sang Nam T, Se Paik K, Leem JW (2004) Periph-eral glutamate receptors contribute to mechanical hyperalgesia ina neuropathic pain model of the rat. Neuroscience 128:169–176.

Jang JH, Kim KH, Nam TS, Lee WT, Park KA, Kim DW, Leem JW(2007) The role of uninjured C-afferents and injured afferents in thegeneration of mechanical hypersensitivity after partial peripheral
nerve injury in the rat. Exp Neurol 204:288–298.


Ji RR, Kohno T, Moore KA, Woolf CJ (2003) Central sensitization andLTP: do pain and memory share similar mechanisms? TrendsNeurosci 26:696–705.

Ji RR, Strichartz G (2004) Cell signaling and the genesis of neuro-pathic pain. Sci STKE 2004:reE14.

Kawasaki Y, Xu ZZ, Wang X, Park JY, Zhuang ZY, Tan PH, Gao YJ,Roy K, Corfas G, Lo EH, Ji RR (2008) Distinct roles of matrixmetalloproteases in the early- and late-phase development of neu-ropathic pain. Nat Med 14:331–336.

Khalil Z, Khodr B (2001) A role for free radicals and nitric oxide indelayed recovery in aged rats with chronic constriction nerve injury.Free Radic Biol Med 31:430–439.

Kim SJ, Song SK, Kim J (2000) Inhibitory effect of nitric oxide onvoltage-dependent calcium currents in rat dorsal root ganglioncells. Biochem Biophys Res Commun 271:509–514.

Kimura S, Watanabe K, Yajiri Y, Motegi T, Masuya Y, Shibuki K,Uchiyama S, Homma T, Takahashi HE (1999) Cerebrospinal fluidnitric oxide metabolites in painful diseases. Neuroreport 10:275–279.

Konigsmark B (1970) Methods for the counting of neurons. In: Con-temporary research methods in neuroanatomy, (Nauta WJH andEbbesson SOE, eds), pp 315–340. New York: Springer-Verlag.

Lee DH, Singh JP, Lodge D (2005) Experiments with nitric oxidesynthase inhibitors in spinal nerve ligated rats provide no evidenceof a role for nitric oxide in neuropathic mechanical allodynia. Neu-rosci Lett 385:179–183.

Luo ZD, Chaplan SR, Scott BP, Cizkova D, Calcutt NA, Yaksh TL(1999) Neuronal nitric oxide synthase mRNA upregulation in ratsensory neurons after spinal nerve ligation: lack of a role in allo-dynia development. J Neurosci 19:9201–9208.

Luo ZD, Cizkova D (2000) The role of nitric oxide in nociception. CurrRev Pain 4:459–466.

Ma W, Eisenach JC (2003) Cyclooxygenase 2 in infiltrating inflamma-tory cells in injured nerve is universally up-regulated followingvarious types of peripheral nerve injury. Neuroscience 121:691–704.

Ma W, Eisenach JC (2007) Neuronal nitric oxide synthase is upregu-lated in a subset of primary sensory afferents after nerve injurywhich are necessary for analgesia from alpha2-adrenoceptor stim-ulation. Brain Res 1127:52–58.

Matsuoka I, Meyer M, Thoenen H (1991) Cell-type-specific regulationof nerve growth factor (NGF) synthesis in non-neuronal cells:comparison of Schwann cells with other cell types. J Neurosci11:3165–3177.

Meller ST, Gebhart GF (1993) Nitric oxide (NO) and nociceptive pro-cessing in the spinal cord. Pain 52:127–136.

Meller ST, Pechman PS, Gebhart GF, Maves TJ (1992) Nitric oxidemediates the thermal hyperalgesia produced in a model of neuro-pathic pain in the rat. Neuroscience 50:7–10.

Milligan ED, Twining C, Chacur M, Biedenkapp J, O’Connor K, PooleS, Tracey K, Martin D, Maier SF, Watkins LR (2003) Spinal glia andproinflammatory cytokines mediate mirror-image neuropathic painin rats. J Neurosci 23:1026–1040.

Murwani R, Hodgkinson S, Armati P (1996) Tumor necrosis factoralpha and interleukin-6 mRNA expression in neonatal Lewis ratSchwann cells and a neonatal rat Schwann cell line followinginterferon gamma stimulation. J Neuroimmunol 71:65–71.

Oh SB, Tran PB, Gillard SE, Hurley RW, Hammond DL, Miller RJ(2001) Chemokines and glycoprotein120 produce pain hypersen-sitivity by directly exciting primary nociceptive neurons. J Neurosci21:5027–5035.

Pan HL, Chen SR, Eisenach JC (1998) Role of spinal NO in antiallo-dynic effect of intrathecal clonidine in neuropathic rats. Anesthesi-
ology 89:1518–1523.
Prast H, Philippu A (2001) Nitric oxide as modulator of neuronalfunction. Prog Neurobiol 64:51–68.

Qian Y, Chao DS, Santillano DR, Cornwell TL, Nairn AC, Greengard P,Lincoln TM, Bredt DS (1996) cGMP-dependent protein kinase indorsal root ganglion: relationship with nitric oxide synthase andnociceptive neurons. J Neurosci 16:3130–3138.

Ren K, Dubner R (2010) Interactions between the immune and ner-vous systems in pain. Nat Med 16:1267–1276.

Saito H, Dahlin LB (2008) Expression of ATF3 and axonal outgrowthare impaired after delayed nerve repair. BMC Neurosci 9:88.

Scholz J, Woolf CJ (2007) The neuropathic pain triad: neurons, im-mune cells and glia. Nat Neurosci 10:1361–1368.

Shamash S, Reichert F, Rotshenker S (2002) The cytokine network ofWallerian degeneration: tumor necrosis factor-alpha, interleukin-1alpha, and interleukin-1beta. J Neurosci 22:3052–3060.

Shibasaki M, Sasaki M, Miura M, Mizukoshi K, Ueno H, Hashimoto S,Tanaka Y, Amaya F (2010) Induction of high mobility group box-1in dorsal root ganglion contributes to pain hypersensitivity afterperipheral nerve injury. Pain 149:514–521.

Shortland PJ, Baytug B, Krzyzanowska A, McMahon SB, Priestley JV,Averill S (2006) ATF3 expression in L4 dorsal root ganglion neu-rons after L5 spinal nerve transection. Eur J Neurosci 23:365–373.

Snyder SH (1992) Nitric oxide: first in a new class of neurotransmitters.Science 257:494–496.

Steel JH, Terenghi G, Chung JM, Na HS, Carlton SM, Polak JM (1994)Increased nitric oxide synthase immunoreactivity in rat dorsal rootganglia in a neuropathic pain model. Neurosci Lett 169:81–84.

Takahashi M, Kawaguchi M, Shimada K, Konishi N, Furuya H, Na-kashima T (2004) Cyclooxygenase-2 expression in Schwann cellsand macrophages in the sciatic nerve after single spinal nerveinjury in rats. Neurosci Lett 363:203–206.

Takahashi T, Kondoh T, Ohtani M, Homma H, Fukuda M (1999)Association between arthroscopic diagnosis of temporomandibularjoint osteoarthritis and synovial fluid nitric oxide levels. Oral SurgOral Med Oral Pathol Oral Radiol Endod 88:129–136.

Thacker MA, Clark AK, Bishop T, Grist J, Yip PK, Moon LD, ThompsonSW, Marchand F, McMahon SB (2009) CCL2 is a key mediator ofmicroglia activation in neuropathic pain states. Eur J Pain13:263–272.

Thomas DA, Ren K, Besse D, Ruda MA, Dubner R (1996) Applicationof nitric oxide synthase inhibitor, N omega-nitro-L-arginine methylester, on injured nerve attenuates neuropathy-induced thermalhyperalgesia in rats. Neurosci Lett 210:124–126.

Tsujino H, Kondo E, Fukuoka T, Dai Y, Tokunaga A, Miki K, YonenobuK, Ochi T, Noguchi K (2000) Activating transcription factor 3(ATF3) induction by axotomy in sensory and motoneurons: a novelneuronal marker of nerve injury. Mol Cell Neurosci 15:170–182.

Yoon YW, Sung B, Chung JM (1998) Nitric oxide mediates behavioralsigns of neuropathic pain in an experimental rat model. Neurore-port 9:367–372.

Yoshimura N, Seki S, de Groat WC (2001) Nitric oxide modulatesCa(2�) channels in dorsal root ganglion neurons innervating raturinary bladder. J Neurophysiol 86:304–311.

Zelenka M, Schafers M, Sommer C (2005) Intraneural injection ofinterleukin-1beta and tumor necrosis factor-alpha into rat sciaticnerve at physiological doses induces signs of neuropathic pain.Pain 116:257–263.

Zhuang ZY, Wen YR, Zhang DR, Borsello T, Bonny C, Strichartz GR,Decosterd I, Ji RR (2006) A peptide c-Jun N-terminal kinase (JNK)inhibitor blocks mechanical allodynia after spinal nerve ligation:respective roles of JNK activation in primary sensory neurons andspinal astrocytes for neuropathic pain development and mainte-
nance. J Neurosci 26:3551–3560.
(Accepted 25 May 2011)(Available online 12 June 2011)

upregulation of neuronal nitric oxide synthase in the … of...key words: peripheral nerve injury,...

Documents