N

The Journal of Pain, Vol 13, No 5 (May), 2012: pp 498-506Available online at www.jpain.org and www.sciencedirect.com

(1)-Naloxone, an Opioid-Inactive Toll-Like Receptor 4 Signaling

Inhibitor, ReversesMultipleModels of Chronic Neuropathic Pain in

Rats

Susannah S. Lewis,* Lisa C. Loram,* Mark R. Hutchinson,y Chien-Ming Li,z

Yingning Zhang,* Steven F. Maier,* Yong Huang,z Kenner C. Rice,x and Linda R. Watkins**Department of Psychology & Neuroscience, University of Colorado at Boulder, Boulder, Colorado.yDiscipline of Physiology, School of Medical Sciences, University of Adelaide, Adelaide, South Australia, Australia.zDrug Studies Unit, Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco,California.xChemical Biology Research Branch, National Institute on Drug Abuse and National Institute on Alcohol Abuse andAlcoholism National Institutes of Health, Rockville, Maryland.

Received2012.This worDA02313ported inNIAAA. M2007-201(DP11010Address r& NeuroBoulder,

1526-590

ª 2012 breserved

doi:10.10

498

Abstract: Previous work demonstrated that both the opioid antagonist (-)-naloxone and the non-

opioid (1)-naloxone inhibit toll-like receptor 4 (TLR4) signaling and reverse neuropathic pain ex-

pressed shortly after chronic constriction injury. The present studies reveal that the TLR4 contributes

to neuropathic pain in another major model (spinal nerve ligation) and to long established (2–4

months) neuropathic pain, not just to pain shortly after nerve damage. Additionally, analyses of

plasma levels of (1)-naloxone after subcutaneous administration indicate that (1)-naloxone has com-

parable pharmacokinetics to (-)-naloxone with a relatively short half-life. This finding accounts for

the rapid onset and short duration of allodynia reversal produced by subcutaneous (1)-naloxone.

Given that toll-like receptor 2 (TLR2) has also recently been implicated in neuropathic pain, cell lines

transfected with either TLR4 or TLR2, necessary co-signaling molecules, and a reporter gene were

used to define whether (1)-naloxone effects could be accounted for by actions at TLR2 in addition

to TLR4. (1)-Naloxone inhibited signaling by TLR4 but not TLR2. These studies provide evidence

for broad involvement of TLR4 in neuropathic pain, both early after nerve damage and months later.

Additional, they provide further support for the TLR4 inhibitor (1)-naloxone as a novel candidate for

the treatment of neuropathic pain.

Perspective: These studies demonstrated that (1)-naloxone, a systemically available, blood-brain

barrier permeable, small molecule TLR4 inhibitor can reverse neuropathic pain in rats, even months

after nerve injury. These findings suggest that (1)-naloxone, or similar compounds, be considered

as a candidate novel, first-in-class treatment for neuropathic pain.

ª 2012 by the American Pain Society. Published by Elsevier Inc. All rights reserved

Key words: TLR4, neuropathic pain, chronic constriction injury, spinal nerve ligation, HEK293-TLR4.

europathic pain is difficult to treat, with manypatients receiving insufficient relief. It is increas-ingly recognized that the proinflammatory mole-

June 14, 2011; Revised January 29, 2012; Accepted February 22,

k was funded in part by NIH Grants DA024044, DE017782,2 and NIDA contract N01DA-9-8883. This work was also sup-part by the NIH Intramural Research Programs of NIDA andark R. Hutchinson is a NHMRC CJ Martin Fellow (ID 465423;

0) and an Australian Research Council Research Fellow0297). HEK-TLR4 cells were gifted from Avigen.eprint requests to Susannah S. Lewis, Department of Psychologyscience, Campus Box 345, University of Colorado at Boulder,CO80309-0345. E-mail: susannah.lewis@colorado.edu

0/$36.00

y the American Pain Society. Published by Elsevier Inc. All rights

16/j.jpain.2012.02.005

cules secreted by glial cells following nerve damagecontribute to the maintenance of neuropathic pain.20

However, there are currently no approved treatmentsthat target this response. One mechanism proposedfor the proinflammatory activation of glial cells in re-sponse to nerve injury is activation of toll-like receptor4 (TLR4). TLR4 is a pattern recognition receptor shownto be important in the mediation of neuropathicpain,3,27 likely through TLR4 detection of damage-associated signals.24 TLR4 activation by agonists suchas lipopolysaccharide (LPS) produces robust proinflam-matory responses.14 TLR4 is found on, at least, microgliain the brain2 and signals through a receptor complexincluding CD14 and MD-2, to create proinflammatoryresponses.5

Lewis et al The Journal of Pain 499

TLR4 signaling canbe temporarily inhibited in anumberof ways. Two forms of inactive LPS competitively antago-nize LPS binding, small interfering RNA (siRNA) againstTLR4 can block proinflammatory responses to LPS,28 andthe smallmoleculenaloxone canblockTLR4 signaling.10,11

Notably, inhibition of TLR4 by all 4 methods has beenshown to reduce neuropathic pain expressed shortlyfollowing sciatic chronic constriction injury (CCI).10,28

However, for TLR4 to be a potentially viable drug targetfor the treatment of neuropathic pain, TLR4 mustremain important in the mediation of neuropathic painacross time and be important in neuropathic pain ofvarying etiologies. Currently, it is unknown how longfollowing injury TLR4 continues to play a role inmaintaining chronic pain, nor whether acute TLR4antagonismcan reverseothermodelsofneuropathicpain.The inhibition of TLR4 signaling by naloxone is espe-

cially intriguing from a clinical perspective. While siRNAand the large lipopolysaccharide molecules are unableto cross the blood-brain barrier, both (1)- and (-)-nalox-one have central nervous system effects following sys-temic administration, making it a better candidatetreatment. While the (-)-naloxone isomer antagonizesopioid receptors, both the (-)- and (1)-naloxone isomersinhibit TLR4 signaling and reverse neuropathic pain fromsciatic CCI.10 (1)-Naloxone, then, is a blood-brain barrierpermeable small molecule that inhibits TLR4 receptorsignaling without antagonizing classical opioid recep-tors. (1)-Naloxone has also been screened on a widerange of central nervous system targets with no knownoff-target effects.22

In order to better characterize the breadth of effect of(1)-naloxone in treating neuropathic pain, we tested(1)-naloxone against allodynia resulting from spinalnerve ligation (SNL), a second neuropathic pain model.Comparisons of CCI and SNL demonstrate a more robustmechanical allodynia in SNL.12 In addition, SNL hasa greater involvement of the sympathetic nervous sys-tem, also found clinically. Therefore, it is useful to dem-onstrate that a novel compound is effective in a second,distinctly different animal model of neuropathic pain.To establish whether TLR4 remains an important media-tor of neuropathic pain across time, (1)-naloxone wastested against CCI months after nerve injury, far longerthan commonly tested in experimental pain models.This work seeks to further our understanding of the

role of TLR4 in neuropathic pain through the study ofnaloxone as a TLR4 inhibitor. Specifically, we hypothe-size that (1)- and (-)-naloxone will have similar pharma-cokinetic profiles and the ability of (1)- and (-)-naloxoneto relieve neuropathic pain will generalize to a secondmajor neuropathic pain model, SNL. Additionally, wepredict that (1)-naloxone will relieve neuropathic painof long duration in both SNL and CCI models, findingswhich would provide the first evidence that TLR4 is im-portant in long-term maintenance of neuropathicpain. Finally, we explore whether (1)-naloxone willalso disrupt toll-like receptor 2 (TLR2) signaling, in addi-tion to TLR4, as TLR2 is a second pattern recognition re-ceptor that has recently been implicated in themediation of neuropathic pain.25

Methods

SubjectsAdult, male Sprague-Dawley rats (Charles River Labo-

ratories, Wilmington, MA) were used in Experiment 1.Adult, male, pathogen-free Sprague-Dawley rats (HarlanLabs, Madison, WI) were used Experiments 2, 3 and 4.Adult, male, pathogen-free Sprague-Dawley rats (sup-plied by Laboratory & Animal Services, Waite Campus,University of Adelaide, Adelaide, AU) were used in Ex-periment 5. Rats were 275 to 300 g at time of injectionin Experiment 1; 130 to 170 g at time of surgery in Exper-iments 2, 3, and 4; and 325 to 375 g at time of surgery inExperiment 5. Rats in all locations were housed intemperature (23 6 3�C) and light (12 h:12 h light:dark)controlled rooms with water and food available ad libi-tum. All habituation and behavioral testing procedureswere performed during the light phase of the daily cycle.Experimental procedure in Experiment 1 was approvedby theUniversity of California, San Francisco InstitutionalAnimal Care and Use Committee. Experimental proce-dures in Experiments 2, 3 and 4 were approved by theUniversity of Colorado, Boulder Institutional AnimalCare and Use Committee. The experimental procedurein Experiment 5 was approved by the University of Ade-laideAnimal Ethics Committee. Each experimental groupcontains 5 to 9 rats with the exception of Experiment 1which contained 3 rats per group.

Drugs(1)-Naloxone was obtained from the National Insti-

tute on Drug Abuse. (-)-Naloxone was obtained fromSigma (St. Louis, MO). For in vivo studies, the toll-like re-ceptor 4 agonist lipopolysaccharide (LPS, Sigma, St.Louis, MO) and the toll-like receptor 2 agonist Pam3CSK4(Pam; Invivogen, San Diego, CA) were used. The vehiclefor all drugs was sterile endotoxin free saline (Hospira,Lake Forest, IL). (1)-Naloxone, (-)-naloxone, Pam and sa-line were confirmed to be endotoxin free at the highestconcentrations used in the present studies by the limulusamebocyte lysate assay (Lonza, Walkersville, MD), per-formed according to the manufacturer’s instructions.

Measurement of (1)-Naloxone Contentin Plasma SamplesA protein precipitation method was used to measure

(1)-naloxone content in collected plasma samples, usinga blinded procedure. An aliquot (20 mL) of rat plasma andadditional 20 mL of 50% acetonitrile were mixed prior tothe precipitation with 75 mL of acetonitrile containingthe isotope-labeled naloxone (naloxone-d5, the internalstandard, 5 ng/mL). The mixture was vortexed for 1 min-ute. After centrifugation (3,000 r.p.m., 10 minutes), thesupernatant was transferred an autosampler vial, andanalyzed by liquid chromatography tandem mass spec-trometry (LC-MS/MS). All analytes were separated ona Synergi Polar-RP column (4.6 � 75 mm, 4 mm; Phenom-enex, Torrance, CA) with a flow rate at 1 mL/minute. Amobile phase system consisting of methanol: water: for-mic acid (28: 72: .2, v/v/v) with 5 mM ammonium formate

500 The Journal of Pain TLR4 Inhibition Reverses Neuropathic Pain

was used to achieve the separation. Mass spectrometricdetection (Micromass Ultima, Beverly,MA)was equippedwith positive ionization by electrospray ionizationand mass scanning by Multiple Reaction Monitoring(m/z 328 / 310 for (1)-naloxone, m/z 333 / 315 fornaloxone-d5) analysis. The standard curve range is 1.00to 400 ng/mL and the lower limit of quantitation is 1.00ng/mL. The LC-MS/MS method was validated with 3inter-day assays (N = 12) and 1 intra-day assay (N = 6);all precision values (coefficient of variation, CV) and ac-curacy values (relative error) were within 15%, suggest-ing that the method was sufficiently reproducible foranalysis of study samples.

Spinal Nerve Ligation (SNL)In Experiments 2, 3 and4, neuropathic painwas induced

through ligation of the fifth and sixth lumbar (L) spinalnerves, as described previously.10,12,13 Briefly, underisoflurane anesthesia, the left L5 and L6 spinal nerveswere isolated and each nerve root tightly ligated with6-0 silk suture. Sham controls had an identical surgeryprocedureexceptwithout the ligationof the spinalnerves.

Chronic Constriction Injury (CCI) of theSciatic NerveA second model of neuropathic pain used in Experi-

ment 5 was the CCI model of partial sciatic nerve injury.1

CCI was performed at the midthigh level of the left hind-leg as previously described.16 In brief, under isofluraneanesthesia, the sciatic nerve was isolated and looselytied with 4 sterile chromic gut sutures (4-0 chromic gut,Ethicon, Somerville, NJ).

Von Frey Test for Tactile AllodyniaAll behavioral testing was conducted by an experi-

menter blind to group assignment. Rats received at least4, 1-hour habituations to the testing apparatus and envi-ronment prior to baseline testing, as in previous stud-ies.15 Baseline response thresholds were determinedprior to nerve injury and rats tested periodically from 3days postinjury until time of drug administration.The response threshold to light touch on both plantar

hind paws was measured using calibrated microfila-ments (von Frey hairs; Stoelting, Wood Dale, IL), as de-scribed previously.17 Briefly, a logarithmic range ofhairs from .406 to 15.136 g force were used, allowingboth analgesia and allodynia to be measured, followingthe standard procedures previously described.4 Re-sponses were fitted to a Gaussian integral psychometricfunction using a maximum-likelihood fitting method asdescribed previously.17

In Vitro Assay for TLR4 SignalingTwo human embryonic kidney-293 (HEK 293) cell lines

were used in Experiment 6. One cell line was stably trans-fectedby Invivogen (SanDiego,CA) tooverexpress humanTLR4 and coreceptor molecules (MD-2, CD14) (referred tohere as HEK-TLR4). The second cell line was stably trans-fected by Invivogen to overexpress human TLR2 and its

coreceptor CD14. In addition, both cell lines stably expressan optimized alkaline phosphatase reporter gene underthe control of a promoter inducible by transcription fac-tors activated as part of the TLR2 and 4 signaling cascades.Secreted alkaline phosphatase (SEAP) protein is producedas a consequence of TLR4 activation.Both HEK-TLR4 and HEK-TLR2 cell lines were grown at

37�C (5%CO2; VWR IncubatorModel 2300; Radnor, PA) in10-cm dishes (Greiner Bio-One, CellStar 632171; Monroe,NC) in normal supplement selectionmedia (DMEMmedia[Invitrogen, Carlsbad, CA] supplemented with 10% fetalbovine serum [Hyclone; Logan, UT], HEK-TLR4 selecton[Invivogen]; Penicillin 10,000 U/mL [Invitrogen]; strepto-mycin 10 mg/mL [Invitrogen], Normocine [Invivogen],and 200 nM L-Glutamine [Invitrogen]). The cells werethen plated for 48 hours in 96 well plates (Microtest 96well plate, flat bottom, Becton Dickinson; 5 � 103 cells/well) with the same media. After 48 hours, supernatantswere removed and replaced with fresh media. Drugstested were added in concentrations indicated and incu-bated for 24 hours. Supernatants (15 ml) were then col-lected from each well for immediate assay.SEAP levels in the supernatants were assayed using the

Phospha-Light System (Applied Biosystems, Foster City,CA) according to the manufacturer’s instructions. Thischemiluminescent assay incorporates Tropix CSPD chemi-luminescent substrate. The 15-ml test samples werediluted in 45 ml of 1� dilution buffer, transferred to96-well plates (Thermo, Walthma, MA), and heated at65�C in a water bath (Model 210, Fisher Scientific, Pitts-burgh, PA) for 30 minutes, then cooled on ice to roomtemperature. Assay buffer (50 ml/well) was added and,5 minutes later, reaction buffer (50 ml/well) is added andallowed to incubate for 20minutes at room temperature.The light output is then measured in a microplate lumin-ometer (Dynex Technologies, #IL213.1191, Chantilly, VA).

StatisticsPharmacokinetic parameters in Experiment 1 were cal-

culated using WinNonlin software (Pharsight Corpora-tion, Mountain View, CA). GraphPad Prism (version 5for Windows, San Diego, CA) software was used for allother statistical analyses. Two-way repeated measuresANOVAs with Bonferroni post hoc tests, when appropri-ate, were used to determine statistical significance be-tween behavioral measures, with the exception of thesham rats in Experiment 2, which were analyzed witha 1-way repeated measures ANOVA. One-way ANOVAswith appropriate Bonferroni post hocswere used to com-pare experimental groups on in vitro experiments and toconfirm that there were no baseline differences on be-havioral measures. For all analyses, P < .05 was consid-ered significant.

Experimental Procedures

Experiment 1: What is the PharmacokineticProfile of Systemic (1)-Naloxone?

All rats had surgically implanted jugular vein cannula(single-side jugular vein cannula). All animals were

Lewis et al The Journal of Pain 501

fasted 16 hours prior to dosing. (1)-Naloxone in sterilephosphate buffered saline at 10 mg/mL was adminis-tered orally (p.o.) and subcutaneously (s.c.) at a doseof 50 and 10 mg/kg, respectively. Blood samples (approx-imately .2 mL) were collected via the jugular vein cathe-ter at .25, .5, 1, 2, 4, and 8 hours for both treatments.Blood samples were centrifuged at approximately10,000 rpm for 5 minutes at 4�C to obtain plasma. Allplasma samples were stored at �20�C until analysis, de-scribed previously.

Experiment 2:What Dose of (-)-Naloxone Is Ef-fective on Neuropathic SNL Pain?

All rats underwent baseline (BL) von Frey testing fol-lowed by SNL surgery. Two weeks following surgery a ro-bust and stable allodynia had developed. Rats weregiven (-)-naloxone doses alternating with saline dosesin a within-subjects design. Group 1 received saline, 10mg/kg s.c. (-)-naloxone, and saline in weeks 2, 3, and 4postsurgery. Group 2 received 1 mg/kg s.c. (-)-naloxone,saline, and 100 mg/kg s.c. (-)-naloxone in weeks 2, 3,and 4 postsurgery. This design was utilized to minimizerat numbers and duration of time spent in a painful con-dition while still providing blinded testing each week.Tactile sensitivity was assessed prior to and 1 and 3 hoursfollowing each subcutaneous injection. These timepoints were selected based on pilot data that showeda maximal effect at 1 hour that had returned to baselineby 3 hours following injection.

Experiment 3: Does (1)-Naloxone ReverseNeuropathic Pain 2 Weeks Following SNL?

All rats underwent baseline (BL) von Frey testing fol-lowed by SNL surgery, as in Experiment 2. Two weeks fol-lowing SNL surgery, after robust and stable tactileallodynia had developed, rats were given 100 mg/kg(1)-naloxone subcutaneously (s.c.) or equivolume salinevehicle in a between-subjects design. An additionalgroup of rats (n = 6) were given sham surgery followedby 100 mg/kg (1)-naloxone s.c. and equivolume salinein a counter-balanced, within-subjects design with a 1-week washout between drugs. For all groups, tactilethresholds were tested prior to and 1 and 3 hours follow-ing s.c. injection.

Experiment 4: Does (1)-Naloxone ReverseNeuropathic Pain at 8 Weeks After SNL?

Eight weeks following SNL surgery, the efficacy of s.c.(1)-naloxone in increasing tactile thresholds was againtested. Six rats from Experiment 3 were used here, alongwith 6 sham rats from another, similar experiment. Allrats underwent BL von Frey testing and then SNL orsham surgery. Beginning 8 weeks following surgery,rats were again assessed for their von Frey responseand were then given (1)-naloxone (100 mg/kg, s.c.) orequivolume saline vehicle in a counterbalanced within-subjects design with a 1-week washout period betweeninjections. Tactile thresholds were assessed prior to and1 and 3 hours following each s.c. injection.

Experiment 5: Does (1)-Naloxone ReverseNeuropathic Pain 16 Weeks After CCI?

All rats underwent BL von Frey testing and then CCIsurgery. Four months later, (1)-naloxone (60 mg/kg,s.c.) or equivolume saline vehicle was given. This lowereddose, relative to preceding experiments, was based bothon pilot testing and necessitated by constraints on (1)-naloxone availability at the time that this study was un-derway. Tactile thresholds were assessed prior to and30, 60, 90, 120, and 180 minutes following s.c. injection.

Experiment 6: Does (1)-Naloxone AntagonizeHEK-TLR4 or HEK-TLR2 Cell SEAP Expression?

HEK-TLR4 cells were stimulated with .1 ng/mL LPS toproduce a robust increase in SEAP signal. The efficacyof 1, 10, and 100 mM (1)-naloxone to inhibit the LPS sig-nal was tested by coincubating each (1)-naloxone dosewith .1 ng/mL LPS. Following a 24-hour coincubation, su-pernatants were then assayed for SEAP expression.Similarly, HEK-TLR2 cells were stimulated with 1 ng/mL

Pam to produce a robust increase in SEAP signal of a sim-ilar magnitude to the LPS signal investigated in the HEK-TLR4 cells. The efficacy of 1, 10 and 100 mM (1)-naloxoneto inhibit the Pam signal was tested by coincubating each(1)-naloxone dose with 1 ng/mL Pam. Following a 24-hour coincubation, supernatants were then assayed forSEAP expression.

Results

Experiment 1: (1)-Naloxone Has a BriefHalf-Life in Blood in Both Oral andSubcutaneous Routes of Administration(-)-Naloxone is known to have a brief half-life in blood

of approximately .5 to .67 hours.9,21 As it is unknownwhether the pharmacokinetics of (1)-naloxone wouldbe the same, the time course of blood levels of (1)-naloxone was assessed in rats after s.c. administration soto guide the route of (1)-naloxone administration andappropriate time points to test for reversal of allodyniain the following studies. The pharmacokinetics of (1)-naloxone was assessed in rats after oral (50 mg/kg) ands.c. (10 mg/kg) administration. A higher oral dose wasused, as a low absorption would be anticipated if (1)-naloxone were to prove to be similar to (-)-naloxone inthis regard. As shown in Fig 1, (1) naloxonewas absorbedvery rapidly after p.o. and s.c. administration. The half-lifevalues (mean6 standard deviation) of (1) naloxonewere2.33 6 1.08 hours and 1.57 6 .784 hours for p.o. and s.c.administration, respectively. The AUC0/N (area underthe curve) ofp.o. (50mg/kg) and s.c. (10mg/kgadministra-tion) were 82.7 and 1,825 hours *ng/mL, respectively. Therelative bioavailability of (1)-naloxone (p.o. versus s.c.)was only .9%. The loworal bioavailability of (1)-naloxonein this study was consistent with (-)-naloxone in the litera-ture (<1% oral bioavailability).9

Given this pharmacokinetic profile, behavioral mea-sures were assessed in subsequent studies at 1 and 3

Figure 1. (1)-Naloxone has a short plasma half-lifewhen givens.c. (10 mg/kg) or p.o. (50 mg/kg) in rats (n = 3 per group). Thehalf-life values of (1)-naloxone were 2.33 6 1.08 hours and1.57 6 .784 hours for p.o. and s.c. administration. The AUC0/N

(area under the curve) of p.o. (50 mg/kg) and s.c. (10 mg/kg ad-ministration) were 82.7 and 1,825 hr*ng/mL, respectively. Therelative bioavailability of (1)-naloxone (p.o. versus s.c.) wasonly .9%. The data are presented as mean 6 SD.

Figure 2. (-)-Naloxone significantly reversed SNL pain at a doseof 100 mg/kg, but not at lower doses. Data shown are from 1hour following s.c. injection as all rats at each dose had returnedto baseline 3 hours following s.c. (-)-naloxone injection. Therewere no baseline differences between groups and all rats devel-oped significant allodynia in this within-subjects design (n = 12).A star (*) indicates a significant (P < .05) between (-)-naloxoneand saline treatments, as indicated by Bonferroni post hoc tests.

502 The Journal of Pain TLR4 Inhibition Reverses Neuropathic Pain

hours following subcutaneous administration, with theprediction of greater drug efficacy at 1 hour.

Experiment 2: (-)-Naloxone ReversesNeuropathic Pain at 100mg/kg but Not at10 or 1 mg/kgFor all in vivo behavioral studies, no differences in tac-

tile thresholds were detected between the ipsilateraland contralateral paws prior to or following drug admin-istration; thus, results here and subsequent behavioralstudies are reported as the average of both paws.Hutchinson et al10 demonstrated that a single 100-mg/

kg (s.c.) dose of either (1)- or (-)-naloxone reversed neu-ropathic pain induced by CCI. To definewhether TLR4 ac-tivation may contribute to neuropathic pain beyond justthat from CCI, the effects of naloxone were tested ona second major neuropathic pain model. SNL was chosenfor study here as it has been reported to be amore robustinducer of mechanical allodynia,12 less inflammatory, aswell as being mediated by mechanisms distinct fromthose of CCI. Here we assessed whether naloxone alsowas effective in SNL. (1)-Naloxone is not commerciallyavailable and in very limited supply from NIDA. Givenavailability constraints of (1)-naloxone and the previ-ously demonstrated parity between (1)- and (-)-nalox-one efficacy on in vitro,10 in vivo,10 and in silico11 testson blockade of TLR4, the effective dose of (-)-naloxone,rather than (1)-naloxone, on SNL-induced pain was firstdetermined as a prediction ofwhat (1)-naloxone dose totest in Experiment 3.The dose-response function for the (-)-naloxone rever-

sal of SNL pain was determined using a within-subjectsdesign that took place between 2 to 4weeks postsurgery,as described in methods. Rats were habituated and base-line testing conducted as described in themethods. Base-lines are lower than typically reported for CCI due to theparticularly young and small rats used in the SNL model(�150 g at time of surgery) and were consistent acrossExperiments 2, 3, and 4. All rats developed allodynia

following SNL. There was a significant interaction be-tween treatment group and pain thresholds (F(9, 9) =9.37, P< .05, Fig 2), andpost hoc tests showed (-)-naloxonesignificantly reversedallodyniaat the100-mg/kgdose, butnotat the10-or1-mg/kgdoses, 1hour following injection.In all cases, rats were at stable allodynia levels prior to(-)-naloxone administration and returned to baselinethresholds by the 3-hour time point. This result indicatedthat the effect of (-)-naloxone generalized beyond neuro-pathic pain from CCI to a second major neuropathic painmodel, and that the most effective dose of those testedfor reversal of SNL-induced neuropathic pain by (1)-naloxone in Experiment 3 was 100 mg/kg.

Experiment 3: (1)-Naloxone ReversesSNL-Induced Neuropathic PainExperiment 2 showed that 100 mg/kg (-)-naloxone s.c.

was sufficient to reverse SNL-induced neuropathic pain.Prior research has shown that the TLR4 antagonists (1)-naloxone and (-)-naloxone were effective at reversingCCI-induced neuropathic pain10 at the same dose. There-fore, 100 mg/kg (1)-naloxone was selected for subse-quent studies of TLR4 mediation of SNL-inducedneuropathic pain. Testing (1)-naloxone, rather thanjust (-)-naloxone, is important given that the effects ofthe 2 stereoisomers are distinct in that (1)-naloxone failsto bind classical opioid receptors.Whether (1)-naloxone produced a similar reversal of

SNL-induced neuropathic pain as (-)-naloxone producedin Experiment 2 was tested here. Baseline tactile sensitiv-itywasmeasured followedbySNLor shamsurgery.All ratsthat underwent SNL surgery developed significant allody-nia by 14 days after SNL, while sham rats failed to developallodynia. There were no significant differences betweengroups in baseline tactile sensitivity (F(2,16) = 1.21, P > .05).ShamandSNL ratswere run inat separate times and indif-ferent experimental designs and are thus analyzed sepa-rately to determine the effect of (1)-naloxone on eachsurgery. Rats that received SNL surgery did not differ in

Figure 3. (1)-Naloxone reverses SNL pain 2weeks following in-jury. There were no baseline differences between groups. Allrats developed significant allodynia following SNL surgerywhich was significantly reversed 1 hour, but not 3 hours, follow-ing 100 mg/kg s.c. (1)-naloxone. (1)-Naloxone had no effect onresponse thresholds in sham operated rats, compared to saline(n = 6). A star (*) indicates a significant difference (P < .05) be-tween (1)-naloxone- (n = 7) and saline- (n = 6) treated SNLrats, as indicated by Bonferroni post hoc tests.

Figure 4. (1)-Naloxone reverses SNL pain 8weeks following in-jury. Rats with SNL surgery (n = 9) developed significant allody-nia while rats which underwent a sham surgery (n = 5) did not.SNL-induced allodynia was significantly reversed 1 hour follow-ing a 100-mg/kg s.c. dose of (1)-naloxone, but saline vehicle didnot significantly change SNL allodynia in this within-subjects de-sign. Neither (1)-naloxone nor saline altered the responsethresholds of the rats who received sham surgery. A star (*) indi-cates a significant difference (P < .05) between (1)-naloxoneand saline treatments in the SNL rats.

Lewis et al The Journal of Pain 503

the established allodynia prior to injection (t11 = . 56, P >.05). In the SNL groups, 100 mg/kg (1)-naloxone signifi-cantly reversed the allodynia compared to saline controls(F(3, 11) = 5.80, P < .05, Fig 3), with post hoc comparisonsshowing significant reversal at 1 hour after injection (t11= 3.83, P < .05), but not at 3 hours (t11 = .09; P > .05) . Inthe sham rats, there was no effect of (1)-naloxone onvon Frey behaviors (F(6,5) = 1.20, P > .05).Taken together, the results of Experiments 2 and 3

support that blockade of TLR4 by both (-)-naloxoneand (1)-naloxone, at a dose of 100 mg/kg s.c., was suffi-cient to reverse established SNL-induced neuropathicpain 2 weeks following nerve injury while failing to alterpain thresholds in sham rats. Additionally, the magni-tude of (1)- and (-)-naloxone reversal of SNL-induced al-lodynia was comparable, providing further evidence forthe nonstereoselective nature of naloxone reversal ofneuropathic pain reported previously10 and suggestivethat endogenous opioids are not tempering the expres-sion of SNL allodynia, as one would have anticipateda pronociceptive effect of (-)-naloxone actions via classi-cal opioid receptor binding had that been the case.

Experiment 4: (1)-Naloxone ReversesLong-Established SNL-InducedNeuropathic PainLong-established neuropathic pain may involve differ-

ent underlying mechanisms than neuropathic pain ex-pressed shortly after injury. To test whether TLR4remains important for the mediation of long-established neuropathic pain, rats 8 weeks followingSNL or sham surgery were given (1)-naloxone (100 mg/kg, s.c.) or equivolume saline vehicle in a within-subjects design. There were no significant differencesin baseline thresholds (t12 = 1.8, P > .05). (1)-Naloxonesignificantly reversed established allodynia in rats with

SNL 1 hour following injection, but did not alter the tac-tile sensitivity of rats that underwent the sham surgery(F(5,12) = 19.2 P < .05, Fig 4). This finding is the first to dem-onstrate that TLR4 remains importantly involved in themaintenance of long-established neuropathic pain. Ad-ditionally, this experiment confirmed that TLR4 blockadehad no effect on tactile thresholds of rats that under-went sham surgery 8 weeks prior to testing when com-pared to sham plus vehicle.

Experiment 5: (1)-Naloxone ReversesLong-Established CCI-InducedNeuropathic PainExperiment 4 showed that (1)-naloxone was capable

of reversing long-established SNL pain. To determine ifthis ability to reverse long-established neuropathic paingeneralized to other models, and to test this ability inneuropathic pain of even longer duration,18,19 rats 4months post-CCI surgery were given (1)-naloxone (60mg/kg) or equivolume saline vehicle. As noted above,a 60-mg/kg dose was used here due to severely limited(1)-naloxone availability while this study was underway.There were no baseline differences between groups (t10= .05 P > .05) and no difference postoperatively prior toinjection. (1)-Naloxone caused a significant reversal ofCCI-induced allodynia (F(3, 10) = 26.1, P < .05. Fig 5), Posthoc tests showed a significant reversal at 1 hour, butnot 3 hours, following injection.

Experiment 6: (1)-NaloxoneAntagonizes Stimulated HEK-TLR4 CellSEAP Expression, but Not StimulatedHEK-TLR2 Cell SEAP ExpressionWhile TLR4 has been the TLR most studied in prior

studies of neuropathic pain, it is now clear that it is not

Figure 5. (1)-Naloxone reversed CCI pain of 4 months dura-tion. There were no significant differences at baseline testingbetween groups. All rats developed chronic allodynia followingCCI surgery. (1)-Naloxone (60 mg/kg, s.c., n = 6) significantly re-versed allodynia at 1 hour, but not 3 hours, following injectioncompared to saline injected rats (n = 6). A star (*) indicates a sig-nificant difference (P < .05) between (1)-naloxone- and saline-treated rats.

Figure 6. (1)-Naloxone antagonizes stimulated HEK-TLR4 cellSEAP expression, but not stimulated HEK-TLR2 cell SEAP expres-sion. The increase in SEAP expression seen in LPS stimulatedHEK-TLR4 cells was significantly blocked by coincubation with1, 10, or 100 mM (1)-naloxone. The increase in SEAP expressioncaused by Pam inHEK-TLR2 cells was unaffected by coincubationwith 1, 10, or 100 mM (1)-naloxone. A star (*) indicates a signifi-cant decrease (P < .05) from vehicle/agonist (LPS [A] or Pam [B])group on Bonferroni post hoc tests.

504 The Journal of Pain TLR4 Inhibition Reverses Neuropathic Pain

the only TLR involved. TLR2 has also recently been recog-nized as a key TLR for the recognition of endogenousdanger signals released by stressed, damaged, and dyingcells. This led to TLR2 being explored for its potential rel-evance to neuropathic pain. TLR2 does indeed appear tobe involved in the mediation of neuropathic pain, giventhat mechanical allodynia is reduced in TLR2 knock-outmice.25While previous evidence supports that (1)-nalox-one can inhibit TLR4 signaling induced by the classicalTLR4 agonist LPS,10 these data do not negate a potentialaction of (1)-naloxone at TLR2 as well. TLR2, in fact,shares cosignaling molecules with TLR4, includingCD147 and sometimes MD-2,7 as well as a similar down-stream intracellular cascade23 including TIRAP.29 Becausethe exact site of (1)-naloxone action along the TLR4 sig-naling cascade is not yet known, (1)-naloxone has thepotential to inhibit TLR2 function as well, given sharedcell signaling steps. If this were true, this would have im-portant implications for how (1)-naloxone effects couldbe interpreted. To investigate whether (1)-naloxone isa TLR2 inhibitor in addition to being a TLR4 inhibitor,HEK-TLR4 and HEK-TLR2 cells were stimulated with anLPS or Pam, respectively, and the ability of (1)-naloxoneto inhibit the agonist response measured.In HEK-TLR4 cells (Fig 6A), (1)-naloxone significantly

inhibited TLR4 signaling in response to LPS (F(5, 10) =8.04, P < .05). Post hoc tests showed that when 1, 10,and 100 mM (1)-naloxone was coincubated with LPS,the SEAP reporter proteins expression was significantlyreduced compared to LPS alone (1 uM: (t10 = 4.38, P <.05); 10 uM: (t10 = 4.83, P < .05); 100 uM: (t10 = 5.51 P <.05)), while there were no significant differences be-tween the 3 doses of (1)-naloxone used (each P > .05)(Fig 6A). In HEK-TLR2 cells (Fig 6B), therewas also a signif-icant difference between groups (F(5,21) = 24.1, P < .05).However, post hoc tests determined that neither 1, 10,nor 100 mM (1)-naloxone coincubated with Pamwas sig-nificantly different from Pam alone (each P > .05) (Fig

6B). Pam-stimulated SEAP expression was significantlyhigher than vehicle (t21 = 6.87, P < .05) or 100 mM (1)-nal-oxone alone (t21 = 5.77, P < .05). In summary, (1)-nalox-one was able to block the HEK-TLR4 effects of LPS, butnot the HEK-TLR2 effects of Pam.

DiscussionThe studies presented here expand our knowledge of

the TLR4 inhibitors (1)- and (-)-naloxone and their abilityto reverse neuropathic pain behaviors. The ability ofTLR4 inhibition by (1)- and (-)-naloxone to reverse neu-ropathic pain in models beyond just short-durationchronic constriction injury (CCI) was tested using an SNLmodel of neuropathic pain. A dose-response studyshowed (-)-naloxone was able to reverse SNL-inducedneuropathic pain at a dose of 100 mg/kg, but not 10mg/kg or 1 mg/kg, 1 hour following s.c. injection. Theopioid-inactive isomer, (1)-naloxone, was similarly effec-tive at this dose on SNL-induced neuropathic pain, fur-ther evidence that that pain reversal by naloxone isnonstereoselective. This behavioral profile of efficacy at1 hour, but not 3 hours, is parsimoniouswith the pharma-cokinetics of (1)-naloxone defined here; namely, that s.c.(1)-naloxone in rat has a half-life of 1.57 6 .784 hour.Our results with SNL provide the first evidence for TLR4mediation of a second major neuropathic pain modelmechanistically distinct from CCI and further supportthat naloxone can reverse neuropathic pain in a nonster-eoselective manner. Furthermore, we present the first

Lewis et al The Journal of Pain 505

evidence that TLR4 is involved not only in newly ex-pressed neuropathic pain but, importantly, continues tobe important in the mediation of even long-standingneuropathic pain. (1)-Naloxone reversed the allodyniaexpressed either 2 months following SNL or 4 months fol-lowing CCI. This supports TLR4 as a potential therapeutictarget given its enduringmediation of these 2major neu-ropathic pain models. Finally, utilizing HEK cells stablytransfected to express either TLR4 or TLR2, the ability of(1)-naloxone to inhibit TLR4 and TLR2 signaling wastested, given that: 1) TLR2 has recently been implicatedin neuropathic pain25; and 2) TLR2 and TLR4 are nowknown to share many elements of their signaling path-ways,7,23,29 raising the potential that (1)-naloxone maypotentially disrupt signaling from both. If so, such anaction would importantly alter interpretation of (1)-naloxone effects. Here, (1)-naloxone inhibited LPS-induced increases in TLR4 reporter protein, as we havepublished previously,10 but did not inhibit the TLR2 ago-nist Pam increase in TLR2 reporter protein, providing thefirst evidence that the in vivo results with (1)-naloxone inCCI and SNL can be appropriately interpreted as reflect-ing involvement of TLR4 rather than TLR2 as well as add-ing to the growing lines of evidence of (1)-naloxone’sspecificity as a TLR4 antagonist given its lack of non-TLR4 actions at other receptors, second messengers, orenzymes.The effectiveness of (1)-naloxone in reversing neuro-

pathic pain up to 4 months after nerve injury providesadditional support for a role of TLR4 in perpetuatingneuropathic pain. Previous work has shown that bothTLR4 knockout and coreceptor CD14 knockout mice de-veloped reduced analgesia and hyperalgesia up to 14days following L5 spinal nerve transection.3,27

Additionally, siRNA blockade of TLR4 given at time ofsurgery was also able to reduce the development of CCIneuropathic pain, up to 14 days following nerveinjury.28 (1)-Naloxone, and the LPS-based TLR4 antago-nists mutant LPS and LPS-RS, were able to reverse neuro-pathic pain 14 days following nerve injury, as shown hereand in Hutchinson et al,10 suggestive that TLR4 is beingcontinually stimulated during these neuropathic painconditions and contributing to the ongoing pain at thistime point. The studies here provide new evidence thatTLR4 continues to contribute to neuropathic pain formonths after injury, far beyond the 14-day period testedin previous work.Given the role of TLR4 in ongoing neuropathic pain, it

is a potential target for novel pharmacological treat-ment. TLR4 is principally found on microglia in the cen-tral nervous system,2 but has been shown to beupregulated on neurons following ischemic injury.26

TLR4 activation induces the production and release ofproinflammatory cytokines, including interleukin-1band tumor necrosis factor-a.5 (1)-Naloxone can reducethe production of proinflammatory cytokines in vitroand reduce the microglial activation marker CD11bin vivo.10 This finding suggests that glial activation inhi-bition and reduced proinflammatory cytokine produc-tion are potential mechanisms for (1)-naloxonereversal of neuropathic pain.

(1)-Naloxone provides a blood-brain barrier perme-able, small molecule, TLR4 inhibitor that reverses neu-ropathic pain in multiple models, months after onset,making it an intriguing potential treatment. While(-)-naloxone is similarly effective on neuropathic pain(here and10), and in silico ligand-receptor modeling in-dicates that (-)-naloxone docks similarly to the TLR4-MD2 complex as does (1)-naloxone.10 (-)-naloxonehas the disadvantage of also inhibiting the endoge-nous opioid system, which reduces its applicability.One limitation of (1)-naloxone is its short duration ofeffect, having a half-life in blood of 1.57 hours afters.c. dose in the rat. In keeping with this pharmacoki-netic profile, the behavioral efficacy of s.c. (1)-nalox-one was apparent at 1 hour but gone by 3 hours hereand in prior studies.10 One potential candidate to im-prove upon the pain reversal of (1)-naloxone is (1)-naltrexone. (1)-Naltrexone is a small molecule that,similar to (1)-naloxone, is the non-opioid active isomerof the opioid receptor antagonist (-)-naltrexone. In hu-mans, (-)-naltrexone has a half life of about 4 to 6hours, significantly longer half-life than the 1 to 1.5hour half-life of naloxone, and is also blood-brainbarrier permeable.6 (1)-Naltrexone inhibits the LPS-induced increase in HEK-TLR4 cell reporter gene ex-pression, reverses neuropathic pain behaviors whengiven intrathecally10, and docks to the TLR4 co-receptor MD-2 in silico.11 However, the lack of commer-cial availability is presently a serious impediment toinvestigating the effects of both (1)-naloxone and(1)-naltrexone.Current evidence suggests that (1)-naloxone is rela-

tively specific to TLR4 receptor activity. In a screen ofcentral nervous system targets other than TLR4, includ-ing neuronal neurotransmitter receptors, neurotrans-mitter transporters, hormone receptors, and enzymes,(1)-naloxone did not have a significant binding affinityto any target.22 As TLR2 has recently been implicated inneuropathic pain and has overlapping signaling cas-cades with TLR4, (1)-naloxone was tested for potentialcross-reactivity to TLR2 as actions here would impact in-terpretation of (1)-naloxone effects. (1)-Naloxone didnot affect TLR2 activation by an agonist, Pam. However,the HEK-TLR2 cell line was not transfected with MD-2,while the HEK-TLR4 cell line was. While MD-2 is not re-quired for TLR2 signaling, it has been shown to en-hance TLR2 responses to the classic TLR4 ligand LPS,among other compounds, while having no effect onsignaling following other TLR2 agonists.8 It remainspossible that (1)-naloxone may effect TLR2 signalingin situations where MD-2 is involved. A lack of off-target effects decreases the chances that (1)-naloxonewill have negative side effects. Additionally, (-)-nalox-one has been used clinically for years to reverse opioidactions, such as respiratory depression, with a goodsafety profile. Given that chronic neuropathic pain isa widespread condition that often remains resistant tocurrent treatments, the effectiveness of (1)-naloxoneon neuropathic pain and ease of administration makeit and like compounds intriguing compounds for futurestudy.

506 The Journal of Pain TLR4 Inhibition Reverses Neuropathic Pain

References

1. Bennett GJ, Xie YK: A peripheral mononeuropathy in ratthat produces disorders of pain sensation like those seen inman. Pain 33:87-107, 1988

2. Bsibsi M, Ravid R, Gveric D, van Noort JM: Broad expres-sion of Toll-like receptors in the human central nervous sys-tem. J Neuropathol Exp Neurol 61:1013-1021, 2002

3. Cao L, Tanga FY, Deleo JA: The contributing role of CD14in toll-like receptor 4 dependent neuropathic pain. Neuro-science 158:896-903, 2009

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

5. Dauphinee SM, Karsan A: Lipopolysaccharide signaling inendothelial cells. Lab Invest 86:9-22, 2006

6. Dawson A: Naloxone, naltrexone and nalmefene, inDart R (ed):Medical Toxicology. Philadelphia, PA, Lippincott,Williams and Wilkins, 2004, pp 228-230

7. Dziarski R, Gupta D: Role of MD-2 in TLR2- and TLR4-mediated recognition of Gram-negative and Gram-positivebacteria and activation of chemokine genes. J EndotoxinRes 6:401-405, 2000

8. Dziarski R, Wang Q, Miyake K, Kirschning CJ, Gupta D:MD-2 enables Toll-like receptor 2 (TLR2)-mediated responsesto lipopolysaccharide and enhances TLR2-mediated re-sponses to Gram-positive and Gram-negative bacteria andtheir cell wall components. J Immunol 166:1938-1944, 2001

9. Hussain MA, Aungst BJ, Kearney A, Shefter E: Buccal andoral bioavailability of naloxone and naltrexone in rats. Inter-national Journal of Pharmaceutics 36:127, 1987

10. Hutchinson MR, Zhang Y, Brown K, Coats BD,Shridhar M, Sholar PW, Patel SJ, Crysdale NY, Harrison JA,Maier SF, Rice KC, Watkins LR: Non-stereoselective reversalof neuropathic pain by naloxone and naltrexone: Involve-ment of toll-like receptor 4 (TLR4). Eur J Neurosci 28:20-29,2008

11. Hutchinson MR, Zhang Y, Shridhar M, Evans JH,Buchanan MM, Zhao TX, Slivka PF, Coats BD, Rezvani N,Wieseler J, Hughes TS, Landgraf KE, Chan S, Fong S,Phipps S, Falke JJ, Leinwand LA, Maier SF, Yin H, Rice KC,Watkins LR: Evidence that opioids may have toll-like recep-tor 4 and MD-2 effects. Brain Behav Immun 24:83-95, 2010

12. Kim KJ, Yoon YW, Chung JM: Comparison of three ro-dent neuropathic pain models. Exp Brain Res 113:200-206,1997

13. Kim SH, Chung JM: An experimental model for periph-eral neuropathy produced by segmental spinal nerve liga-tion in the rat. Pain 50:355-363, 1992

14. Lehnardt S, Massillon L, Follett P, Jensen FE, Ratan R,Rosenberg PA, Volpe JJ, Vartanian T: Activation of innate im-munity in the CNS triggers neurodegeneration througha Toll-like receptor 4-dependent pathway. Proc Natl AcadSci U S A 100:8514-8519, 2003

15. Lewis SS, HutchinsonMR, Rezvani N, Loram LC, Zhang Y,Maier SF, Rice KC, Watkins LR: Evidence that intrathecalmorphine-3-glucuronide may cause pain enhancement viatoll-like receptor 4/MD-2 and interleukin-1beta. Neurosci-ence 165:569-583, 2010

16. Milligan ED, Maier SF, Watkins LR: Sciatic inflammatoryneuropathy in the rat: Surgical procedures, induction of in-flammation, and behavioral testing. Methods Mol Med 99:67-89, 2004

17. Milligan ED, O’Connor KA, Nguyen KT, Armstrong CB,Twining C, Gaykema RP, Holguin A, Martin D, Maier SF,Watkins LR: Intrathecal HIV-1 envelope glycoprotein gp120induces enhanced pain states mediated by spinal cord proin-flammatory cytokines. J Neurosci 21:2808-2819, 2001

18. Milligan ED, Sloane EM, Langer SJ, Hughes TS,Jekich BM, Frank MG, Mahoney JH, Levkoff LH, Maier SF,Cruz PE, Flotte TR, Johnson KW, Mahoney MM, Chavez RA,Leinwand LA, Watkins LR: Repeated intrathecal injectionsof plasmidDNAencoding interleukin-10 produce prolongedreversal of neuropathic pain. Pain 126:294-308, 2006

19. Milligan ED, Soderquist RG, Malone SM, Mahoney JH,Hughes TS, Langer SJ, Sloane EM, Maier SF, Leinwand LA,Watkins LR, Mahoney MJ: Intrathecal polymer-based inter-leukin-10 gene delivery for neuropathic pain. Neuron GliaBiol 2:293-308, 2006

20. Milligan ED, Watkins LR: Pathological and protectiveroles of glia in chronic pain. Nat Rev Neurosci 10:23-36, 2009

21. Ngai SH, Berkowitz BA, Yang JC, Hempstead J, Spector S:Pharmacokinetics of naloxone in rats and inman: Basis for itspotency and short duration of action. Anesthesiology 44:398-401, 1976

22. Northcutt AL, Hutchinson MR, Hiranita T, O’Neill CE,Lewis S, Loram L, Kopajtic T, Galer EL, Cochran TA, BlandST, Maier SF, Bachtell RK, Katz JL, Rice KC, Watkins LR: Evi-dence for toll-like receptor-4 (TLR4) involvement in mor-phine and cocaine reward. Proceedings of the Society forNeuroscience Annual Meeting, Washington, DC, 2011

23. O’Neill LA: The interleukin-1 receptor/Toll-like receptorsuperfamily: 10 years of progress. Immunol Rev 226:10-18,2008

24. Osterloh A, BreloerM: Heat shock proteins: Linking dan-ger and pathogen recognition. Med Microbiol Immunol197:1-8, 2008

25. Shi XQ, Zekki H, Zhang J: The role of TLR2 in nerve injury-induced neuropathic pain is essentially mediated throughmacrophages in peripheral inflammatory response. Glia59:231-241, 2011

26. Tang SC, Arumugam TV, Xu X, Cheng A, Mughal MR,Jo DG, Lathia JD, Siler DA, Chigurupati S, Ouyang X,Magnus T, Camandola S, Mattson MP: Pivotal role for neu-ronal Toll-like receptors in ischemic brain injury and func-tional deficits. Proc Natl Acad Sci U S A 104:13798-13803,2007

27. Tanga FY, Nutile-McMenemy N, DeLeo JA: The CNS roleof Toll-like receptor 4 in innate neuroimmunity and painfulneuropathy. Proc Natl Acad Sci U S A 102:5856-5861, 2005

28. Wu FX, Bian JJ, Miao XR, Huang SD, Xu XW, Gong DJ,Sun YM, Lu ZJ, Yu WF: Intrathecal siRNA against Toll-like re-ceptor 4 reduces nociception in a rat model of neuropathicpain. Int J Med Sci 7:251-259, 2010

29. Yamamoto M, Sato S, Mori K, Hoshino K, Takeuchi O,Takeda K, Akira S: Cutting edge: A novel Toll/IL-1 receptordomain-containing adapter that preferentially activatesthe IFN-beta promoter in the Toll-like receptor signaling. JImmunol 169:6668-6672, 2002