The Role of Sodium Channels in Chronic …...in voltage-gated sodium channels may play a role in...

TN

RMG*I†

M‡

Q§

�

#

H*†

R2TsAUl1©d

The Journal of Pain, Vol 7, No 5 (May), Supplement 3, 2006: pp S1-S29Available online at www.sciencedirect.com

he Role of Sodium Channels in Chronic Inflammatory andeuropathic Pain

on Amir,* Charles E. Argoff,† Gary J. Bennett,‡ Theodore R. Cummins,§

arcel E. Durieux,� Peter Gerner,# Michael S. Gold,** Frank Porreca,†† andary R. Strichartz#

Department of Cell and Animal Biology, Institute of Life Sciences, Hebrew University of Jerusalem, Jerusalem,srael.Cohn Pain Management Center, North Shore University Hospital, New York University School of Medicine,anhasset, New York, New York.

Department of Anesthesia, Faculty of Dentistry, and Centre for Research on Pain, McGill University, Montreal,uebec, Canada.

Department of Pharmacology and Toxicology, Indiana University School of Medicine, Indianapolis, Indiana.Departments of Anesthesiology and Neurological Surgery, University of Virginia, Charlottesville, Virginia.Pain Research Center, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’sospital, Harvard University, Boston, Massachusetts.*Department of Biomedical Sciences, University of Maryland, Baltimore Dental School, Baltimore, Maryland.†Department of Pharmacology, University of Arizona Health Sciences Center, Tucson, Arizona.

Abstract: Clinical and experimental data indicate that changes in the expression of voltage-gatedsodium channels play a key role in the pathogenesis of neuropathic pain and that drugs that blockthese channels are potentially therapeutic. Clinical and experimental data also suggest that changesin voltage-gated sodium channels may play a role in inflammatory pain, and here too sodium-channelblockers may have therapeutic potential. The sodium-channel blockers of interest include localanesthetics, used at doses far below those that block nerve impulse propagation, and tricyclicantidepressants, whose analgesic effects may at least partly be due to blockade of sodium channels.Recent data show that local anesthetics may have pain-relieving actions via targets other than sodiumchannels, including neuronal G protein–coupled receptors and binding sites on immune cells. Some ofthese actions occur with nanomolar drug concentrations, and some are detected only with relativelylong-term drug exposure. There are 9 isoforms of the voltage-gated sodium channel �-subunit, andseveral of the isoforms that are implicated in neuropathic and inflammatory pain states are expressedby somatosensory primary afferent neurons but not by skeletal or cardiovascular muscle. Thisrestricted expression raises the possibility that isoform-specific drugs might be analgesic and lackingthe cardiotoxicity and neurotoxicity that limit the use of current sodium-channel blockers.Perspective: Changes in the expression of neuronal voltage-gated sodium channels may play a keyrole in the pathogenesis of both chronic neuropathic and chronic inflammatory pain conditions. Drugsthat block these channels may have therapeutic efficacy with doses that are far below those thatimpair nerve impulse propagation or cardiovascular function.

© 2006 by the American Pain SocietyKey words: Allodynia, hyperalgesia, inflammatory pain, local anesthetics, neuropathic pain, sodiumchannels.

eceived June 18, 2005; Revised January 13, 2006; Accepted January 20,006.his review is the result of a conference that was supported via an Unre-tricted Educational Grant from Endo Pharmaceuticals.ddress reprint requests to Gary J. Bennett, PhD, Anesthesia Researchnit (McIntyre 1202), McGill University, 3655 Promenade Sir William Os-

er, Montreal, Quebec, Canada H3G 1Y6. E-mail: [email protected]/$32.002006 by the American Pain Society

oi:10.1016/j.jpain.2006.01.444

S1

Dateptiocssdcc

tpvtg

bwaciteontchelgtpstateenpmfntvmtartf

tr

sphtbdb

Ti

mcTrv

S

tltvttia

((taPddtltwtcbptptid

smcpfio

S2 Sodium Channels and Pain

rugs that block voltage-gated sodium channelshave been a mainstay of pain medicine sinceKoller and Freud introduced the use of cocaine as

n ophthalmic anesthetic 120 years ago.208 The introduc-ion of procaine, an ester local anesthetic that could beasily sterilized, the recognition that anesthesia could berolonged significantly if local vasoconstriction was es-ablished with coadministration of epinephrine, and thentroduction of agents with medium and long durationsf action and reduced toxicity (lidocaine and bupiva-aine, respectively), greatly expanded the usefulness ofodium-channel blockade. Today, the administration ofodium-channel blockers with topical, regional, epi-ural, or intrathecal technique is used not only for theontrol of surgical pain but also for the management ofhronic pain conditions.The use of local anesthetics is so familiar that it is easy

o lose sight of the fact that the ability of these drugs torevent nerve impulse propagation is achieved only withery high local concentrations (.1-1.0 mmol/L)—concen-rations that would far exceed the lethal threshold ifiven systemically (plasma levels of 40-60 �mol/L).The possibility that relatively low, systemically tolera-le doses of sodium-channel blockers might be usefulas first made clear by the discovery that procainamidend lidocaine suppress cardiac arrhythmias and that lido-aine and other local anesthetics have antiseizure activ-ty.56 For lidocaine, these therapeutic actions are ob-ained with plasma levels of 5-10 �mol/L. Several lines ofvidence showing that there are useful low-dose actionsf sodium-channel blockers arose from attempts to treateuropathic pain. Early evidence suggested analgesic ac-ivity with orally available lidocaine congeners like to-ainide and mexiletine (for review, see Kalso et al121). Itas long been known that patients with painful periph-ral neuropathy sometimes receive weeks of relief fol-owing a single local anesthetic block of the painful re-ion.12,26,149 Such a long period of pain relief from drugshat were known to have short half-lives invited the sus-icion that the mechanism of action was a placebo re-ponse or even evidence of psychoneurosis. However,here is now abundant evidence10,41,150,196 that rats withn experimental painful peripheral neuropathy also ob-ain many days of pain relief following single systemicxposures to local anesthetic, and it is therefore appar-nt that the effect is physiologic, although the mecha-ism is still not understood. The intravenous lidocainelasma concentrations that relieve clinical and experi-ental neuropathic pain are in the 5-10 �mol/L range,

ar below the concentrations that are needed to blockerve impulse propagation. There is firm evidence thatricyclic antidepressants (TCAs) relieve neuropathic painia a mechanism that is separate from that by which theyodulate mood,153 and there is now reason to believe

hat their analgesic action may be at least partly medi-ted via sodium-channel blockade (see following). Moreecently, evidence has begun to appear that indicateshat changes in the expression of sodium-channel iso-
orms are involved in chronic inflammatory pain condi- c
ions and that sodium-channel blockers may also have aole in the treatment of these conditions.Here we review the evidence for a role of neuronal

odium channels in the pathogenesis of chronic neuro-athic and inflammatory pain conditions. In addition, weighlight a variety of mechanisms that might contributeo the pain relief that is obtained with sodium-channellockers, including mechanisms engaged by very lowoses that may have nothing to do with sodium-channelinding.

he Voltage-Gated Sodium Channel andts IsoformsVoltage-gated sodium channels are transmembraneolecular pores that open and close in response to

hanges in the local electric field across the membrane.he structural features that support channel function areemarkably similar among all of the known isoforms ofoltage-gated sodium channels.

tructure and FunctionThe sodium channel is composed of 1 very large con-

inuous protein, the �-subunit, and 1 or 2 smaller ancil-ary �-subunits.38 The �-subunit contains all of the fea-ures necessary for a functional ion channel, includingoltage sensor, activation gate, ion pore, and inactiva-ion gate. The �-subunits are important for chaperoninghe �-subunits to the plasma membrane and for stabiliz-ng them there, and they can modulate channel gatingnd pharmacology.The �-subunit (Fig 1) consists of 4 homologous domains

D1-D4), each of which contains 6 transmembrane helicesS1-S6) and a short nonhelical region between S5 and S6,he P-segment, which is thought to line the ion perme-tion pathway which is formed by close apposition of the-segments from each of the 4 domains. The ion-con-ucting pore enables ions of the appropriate charge andiameter (sodium being the most permeant) to passhrough the otherwise relatively impermeable lipid bi-ayer. The narrowest part of the pore appears to be closero the outer surface of the channel, and this is the locushere selectivity among ions occurs, the “selectivity fil-

er.”105,200 The inner opening of the pore that faces theytoplasm appears to have a considerably wider vesti-ule, accommodating large organic cations that cannotass through the narrower selectivity filter but can reachhe cytoplasm by permeation through the lipid bilayer oflasma membranes. The cytoplasmic opening also con-ains regions where drugs bind to parts of the channelnvolved in inactivation, the pore-closing process that isriven by depolarization.Opening of sodium channels from their resting closed

tate involves several steps of conformational change,ost in response to forces put on the channels by the

hanged electric field that accompanies membrane de-olarization. Charged segments of the channel, identi-ed as multiple positive amino acid residues on at least 3f the 4 S4 segments,231 “sense” these electrical field
hanges and move within the membrane (in ways that

aclapbSsdt

ahtnfqplti

itbtmtSi

pmcseiccma

C

aTapnarfbiinceu

smwttdrtcbttamgqwdgbaretsincpeistrl

i�

snpMtaamutDbbmaatma

S3ORIGINAL REPORT/Amir et al

re still debated147), effectively displacing their fixedharges “outward” from a location closer to the intracel-ular to one closer to the extracellular surface. Such volt-ge-dependent “activation gating” is associated with aroportional displacement of charge across the mem-rane, the so-called “gating current.”11b Movement of4 segments is apparently coupled to movement of S6egments (again in ways unknown), subtly changing theimensions of the pore and allowing ions to passhrough this open state of the channel.Open channels preferentially transform to a closed in-

ctivated state, a conformation from which reopening isighly unlikely. However, at negative membrane poten-ials, eg, around the resting potential, inactivated chan-els will transition to the closed state, a conformationrom which they are relatively easy to open upon subse-uent depolarization. The rapid inactivation phase ap-ears to be initiated by the binding of a cytoplasmic

oop, which connects D3 to D4 in the inner vestibule ofhe channel, effectively placing an obstruction over thenner opening.

Many different amphipathic amine molecules, includ-ng local anesthetics and class I antiarrhythmics, inhibithe function of sodium channels. Screens of all 4 domainsy site-directed mutagenesis reveal that major contribu-ors to local anesthetic binding are located on S6 seg-ents that line the channel214,215 and that blocking po-

ency is also sensitive to mutations of the inner ring.200

ome of these residues are also involved in binding thenactivation moiety of the D3-D4 linker loop.

Although the binding site for local anesthetics is in theore and in contact with S4 segments, the gating chargeovement that signals the major voltage-dependent

onformational change during channel activation istrongly suppressed by local anesthetics,34b,125b furthervidence for the conformational coupling between “gat-ng” and pore domains. Although the binding pharma-ophore for local anesthetics is very similar among allhannel isoforms, differences in their activation regionsight indirectly influence the effective potency of local

nesthetic-like blockers.

hannel IsoformsNine genes have been identified which encode volt-

ge-gated sodium channel �-subunits in mammals.93

hey differ in terms of their tissue distribution in thedult and developing organism, their electrophysiologicroperties, their pharmacology, and their response toerve injury and inflammation. Somatosensory primaryfferent neurons (ie, dorsal root ganglion (DRG) neu-ons) in the adult rat and mouse express 6 of the 9 iso-orms; another 2 isoforms are present in the normal em-ryonic DRG (Table 1). Coexpression of multiple isoforms

s the rule23; no DRG neuron is known to express only onesoform. Different functional classes of primary afferenteurons (eg, low-threshold mechanoreceptors and noci-eptors) express different mixtures of isoforms. How-ver, there is no evidence that any particular mixture isnique to any particular type of neuron.
We have long known that sodium channels are es- n
ential for the propagation of nerve impulses, and oneight imagine that a single type of sodium channelould be sufficient for such a purpose. Why then are

here 9 �-subunit isoforms? Impulse propagation is nothe only phenomenon mediated via voltage-gated so-ium channels. In the periphery, the sensory neuron’seceptor terminals express macromolecular complexeshat transduce various forms of energy (eg, mechani-al, thermal, or chemical stimuli) into a graded mem-rane depolarization, the generator (or receptor) po-ential. When the generator potential crosses a certainhreshold, the graded depolarization is converted intostream of nerve impulses. Generally, increasing theagnitude and duration of the generator potential

ives rise to nerve impulse streams of increasing fre-uency and duration. Roughly the opposite occurshen a sensory nerve impulse enters the spinal cordorsal horn: The impulse stream is decomposed into araded depolarization that invades the terminalranches of the sensory axon and depolarizes its syn-ptic boutons (the presynaptic potential), causing neu-otransmitter release. All of these processes—the gen-rator potential, its conversion into an impulse stream,he conduction of the impulse stream along the sen-ory axon into the spinal cord, the spread of depolar-zation to the synaptic boutons, and the release ofeurotransmitter—vary depending on which sodium-hannel �-subunit isoforms are present. It thus seemsossible that the diversity among isoforms reflectsvolutionary pressures to fine-tune these phenomena

n the context of the different functional classes ofensory neurons. Moreover, it has been hypothesizedhat different isoforms may associate with a variety ofeceptor molecules to form specialized macromolecu-ar complexes that regulate nociceptor excitability.218

The multiplicity of isoforms may be just the tip of theceberg. It is known that the genes that encode the-subunits have many alternative splice variants, thatome of these variants appear to be unique to DRGeurons, and that peripheral nerve injury changes theatterns of alternative splice variants observed.177

oreover, it is essential to remember that the func-ions of the �-subunits are modulated by their associ-ted �-subunits. Thus, voltage-gated sodium channelctivity in neuropathic and inflammatory conditionsay arise from injury-evoked changes in the �-sub-nits. Five �-subunits have been identified to date, andhere is evidence that at least 3 of them are present inRG neurons. Avulsion of the dorsal roots in man haseen shown to decrease the expression of �1 and �2ut not � 3 in DRG neurons.39 In animals with experi-ental nerve injury,169b the levels of the � 2-subunit

re increased in the DRG somata of those cells whosexons have been injured, and to a lesser extent also inheir uninjured neighbors. Moreover, the null mutantouse that lacks �2-subunits develops less mechano-

llodynia than the wild type following the spared-
erve injury.169b

EC

paTsbmtc

iwtibsmeccatrohfWsh

atdwnipnfttc

cdrcepcwttcste

smlnttspimWblqcrri

I

osshvb

Fseoc(DmovosaclR


lectrophysiologic Properties of Sodium-hannel IsoformsThe channel isoforms have different electrophysiologicroperties. They differ in their thresholds for openingnd in the length of time that the channel remains open.hey also vary in the amount of time that the channelpends in the inactivated state and in the effect of mem-rane potential on this kinetic parameter. At negativeembrane potentials, the different channel isoforms

ransition from the closed inactivated state to the restinglosed state at different rates.Prior to the molecular identification of the channel

soforms, sodium channels were classified functionallyith respect to gating of the current that flowed

hrough them (eg, low- or high-threshold, fast- or slow-nactivating, etc) in conjunction with their sensitivity tolockade by tetrodotoxin (TTX), with some being veryensitive to TTX and others being relatively resistant. Theolecular and electrophysiologic classifications are not

asily reconciled. Equating particular currents with theirorresponding channel isoforms in vivo is difficult, be-ause channel kinetics can be modified by the �-subunitsnd by a variety of endogenous regulators. However,his difficulty is mostly overcome by experiments that useecombinant �-subunit isoforms expressed in Xenopusocytes and other cells, a less complicated situation thatas provided a wealth of unambiguous detail on theunctional characteristics of the different isoforms.92,93

e can now assign the various kinds of currents (and TTXensitivity) to their respective channel isoforms with aigh degree of confidence in most cases (Fig 2).In the normal animal, DRG neurons express a complex

rray of sodium currents, and they are unusual in thathey express both TTX-sensitive and TTX-resistant so-ium currents.133 The population of large DRG neurons,hich contain mostly low-threshold mechanoreceptiveeurons (“touch receptors”), appears to mostly express

soforms that are TTX sensitive and have fast kinetics. Theopulation of small DRG neurons, which contain mostlyociceptive neurons, expresses a mixture of channel iso-orms with both TTX-sensitive fast kinetics and TTX-resis-ant slow kinetics. Much recent research has focused onhe 2 TTX-resistant channels that are expressed in noci-eptors, Nav1.8 and Nav1.9.

igure 1. Structural features of the Na� channel that determineingle peptide of the Na� channel �-subunit in a plasma membach contain 6 alpha helical segments that span the membranef segments, and these bundles converge to form the functionahannel opening results from the primary movement of the posiC). Fast inactivation of the channel follows the binding to the cy-3 to D-4. Ions travel through an open channel along a pore defiembrane penetration of the 4 extracellular loops of protein cof different amino acids on the channel indicate those residueestibule of the channel (bold X, on S6 segments) and at the intern the P region) and are also known to influence stereoselectchematic cross-section of the channel speculates on the manctivation to open the channel, allowing the entry and departlosed (inactivated) channel has a more intimate association wonger between S6 segments (the former pore). Reprinted with
D (ed): Miller’s Anesthesia. Philadelphia, PA, Elsevier Churchill Livin
Nav1.8 produces a slowly activating and inactivatingurrent which has relatively depolarized voltage depen-ences of activation and steady-state inactivation. TTX-esistant sodium current in DRG neurons is believed to bearried primarily via the Nav1.8 isoform. Nav1.8 is prefer-ntially expressed in nociceptive afferents,67 where it ap-ears to be localized in the cell body,181 peripheral re-eptor terminals,30,31,37,127,232 and central terminalsithin the spinal cord dorsal horn.99,116 In peripheral

erminals, the channel appears to underlie spike initia-ion.30 It is yet to be demonstrated that functional Nav1.8hannels are present in central terminals, but a TTX-re-istant current appears to contribute to the release ofransmitter from the central terminals of primary affer-nts.99

Another type of TTX-resistant current, called the per-istent current, is also predominantly expressed in small,ostly nociceptive, DRG neurons.52,178 This current is

ikely produced by the Nav1.9 isoform.62,65,73 The chan-el has a relatively low threshold for activation, such thathe channel may be active at resting membrane poten-ials,52 and it has an exceedingly slow activation rate,uggesting that it contributes little to action potentialropagation. The channel also has an exceedingly slow

nactivation rate, resulting in a “persistent” current thatay contribute to the generation of burst discharge.104

hen the channel was first identified, it was believed toe restricted to the peripheral nervous system, particu-

arly to small-diameter sensory neurons.65,204 Subse-uent anatomic79 and electrophysiologic73 evidenceonfirms that expression of this channel in sensory neu-ons is largely restricted to nociceptive A�- and C-neu-ons. Recent data suggest that the channel is also presentn the enteric nervous system.186

s There a Cocktail Effect?Most of the currently available data focuses on the rolef individual isoforms; however, the complex mixture ofodium-channel isoforms in DRG neurons raises the pos-ibility that in any particular neuron the mixture mayave emergent properties. There are few in vivo or initro data on this question, but preliminary analyses haveeen performed in silico.104,196

Figure 3 shows results obtained with a computer

anesthetic interactions. (A) The consensus arrangement for the. Four domains with homologous sequences (D-1 through D-4)6). Each domain folds within itself to form 1 cylindrical bundlennel’s quaternary structure (B). Activation gating that leads tocharged S4 segments in response to membrane depolarizationsmic end of the channel of part of the small loop that connects

at its narrowest dimension by the P region formed by the partialcting S5 and S6 in each domain. Intentional directed mutationst are involved with local anesthetic (LA) binding in the innergions of the ion-discriminating “selectivity filter” (bold square,for phasic inhibition (bold circle, also on S6 segments). (C) An which S6 segments, forming a “gate,” may realign duringf a bupivacaine molecule by the “hydrophilic” pathway. The

he LA molecule whose favored pathway for dissociation is noission from Strichartz GR, Berde CB: Local anesthetics, in Miller

localrane(S1-Sl chativelytoplanednnes thaior reivityner iure oith tperm

gstone, 2005, pp 573-603.195

mwknt1Rsac

PI

SC

lai

T

Apr

*

† an et


odel of neuronal firing that includes channel isoformsith only fast- or a mixture of fast- and slow-inactivatinginetic parameters matching those measured in DRGeurons. In response to a single brief (.1 msec) supra-hreshold depolarizing stimulus, the model neuron with00% fast-inactivating channels fires a single impulse.emarkably, changing only 5% of the channels to thelowly-inactivating type yields repetitive discharge of rel-tively long duration. It is far from clear whether our

able 1. Properties of Voltage-Gated Sodium C

CHANNEL �-SUBUNIT

ISOFORM

INACTIVATION RATE

(TTXSENSITIVE/RESISTANT)*

DISTRIBUTION IN NORMA

DRG†

Nav1.1 Fast (S) � 75% of all cells; esp.in large cells

Nav1.2 Fast (S) � Low levels; 40% of a

Nav1.3 Fast (S) � Very low or absent inhigh levels in embryo

Nav1.4 Fast (S) �

Nav1.5 Slow (R) � (embryo only)

Nav1.6 Fast (S) � 87% of all cells, esp.in medium and large

Nav1.7 Fast (S) � All cells

Nav1.8 Slow (R) � 50% all cells

Nav1.9 Very slow(persistent) (R)

� 80% of small cells, racells; ca. 50% of all Anociceptors; 90% ofnociceptors

bbreviations:1,2, no �, increase, decrease, or no change in expression of meripheral neuropathy; SNL, spinal nerve ligation model of Kim and Chung131; hizotomy.

Data from International Union of Pharmacology: http://www.iuphar-db.org/iu

Data from Black et al23; Dib-Hajj et al62; Fang et al73; Kim et al130; Renganath

urrent methods could detect such a small change. c

harmacology of Sodium-Channelsoforms

tate-Dependent Binding of Sodium-hannel BlockersA critical aspect of local anesthetic binding that under-

ies many of their therapeutic actions is state-selectiveffinity. Local anesthetics bind relatively weakly to rest-ng closed channels, more tightly and rapidly to open

nel �-Subunit Isoforms

LT RAT

CHANGE POST–NERVE INJURY

CHANGE POST-INFLAMMATION

levels 2 SNL injured DRG130 No �24

No � DN51

s 2 SNL injured DRG130 No �24

lt, but 1 Axot esp small cells; but no� Rhizot22,217

1 small cells24

1 SNL injured DRG; esp.medium and large cells130

1 CCI small cells63

1 DN small and large cells51

n/a n/a

n/a n/a

levels 1 DN small and large cells51 No �24

DN115 1 esp. small cells24,95

2 Axot small cells; but no �Rhizot191

1 esp. small cells24,95,203

2 CCI small cells63; but no �CCI found by Novakovic etal160; decrease only in axotcells57

no � SNL uninjured DRG90

Translocated to axons in CCI,Axot., SNL uninjuredDRG90,136,160

2 DN small and largecells51,115

larged Ct” C-

2 Axot small cells; but no �Rhizot191

2 CCI small cells63; decreaseonly in axot cells57

1 DN small and large cells51

No �24

and/or protein. Axot, axotomy (nerve transection); DN, diabetic painfulhronic constriction injury model of Bennett and Xie17; Rhizot, dorsal

c/sodium.html.

al179; and Waxman et al.217

han

L ADU

high

ll cell

adu

highcells

re in� an“silen

RNACCI, c

phar-i

hannels, and as tightly but more slowly to inactivated

http://www.iuphar-db.org/iuphar-ic/sodium.html

cptasactqettbcagMsiapbtifspc

oirontdr

P

pwttaaogivosatmmld2rnNt

sbrEtgll

FtprcfCrp

Favta


hannels.28,45 As a consequence of these differences,atterns of membrane potential that present more ofhe high-affinity states result in a larger effective block-de.28,188 Two such patterns are germane to the discus-ion of pain: the trains of ectopic impulses that occurfter peripheral nerve injury and the spontaneous dis-harge of sensitized nociceptors innervating inflamedissue. A train of rapidly firing impulses presents a se-uence of open channels that increasingly bind local an-sthetic, to the point that the cumulative binding leavesoo few unbound channels available to support the ac-ion potential.169 The scenario for such use-dependentlock depends on the concentration and binding rateonstants for the drug, the frequency of the impulsesnd, often overlooked, the margin of safety for impulseeneration and propagation in the zone of firing.196

oreover, the prolonged depolarizations produced byome channel isoforms yield a relatively large number ofnactivated channels that support a slow binding of localnesthetics. When the long depolarization is due to aersistently open population of channels, the blockadey the local anesthetic will shorten the depolariza-ion.129 This situation might have particular significancen the case of sodium channels expressed on primary af-erent terminals in the spinal cord. A high density oflowly inactivating sodium channels would support arolonged depolarization, allowing voltage-gated cal-

igure 2. Primary afferent somatosensory neurons express mul-iple voltage-gated sodium channels with distinct voltage de-endence and kinetic properties. (A) Slow TTX-resistant cur-ents. (B) Persistent TTX-resistant currents. (C) Fast TTX-sensitiveurrents. (D) Comparison of voltage dependence of activationor the currents shown in A (solid squares), B (solid circles), and(open circles). All currents were recorded from small-diameter

at DRG neurons. The currents in A and B were recorded in theresence of 500 nmol/L TTX.

ium channels on the terminal’s membrane to remain u

pen longer, which would allow a proportionately largernflux of calcium into the terminal and thereby result in aelative increase in neurotransmitter release. Blockadef only a fraction of the persistently open sodium chan-els, and of more of those channels that are in the inac-ivated state, could dramatically shorten the duration ofepolarization, and this would profoundly reduce theelease of neurotransmitter.

harmacologic Diversity Among IsoformsThe structural diversity among the various isoforms,articularly in those regions of the molecule not involvedith voltage sensing and pore formation, suggests that

hey may have different pharmacologic profiles. Rela-ively little data is available on this question. We havelready noted that there is differential binding for TTXmong the various isoforms. Natural toxins from the ven-ms of poisonous toads, spiders, marine mollusks of theenus Conus (cone snails), dinoflagellates, and other an-

mals also discriminate among isoforms.25,163 The obser-ation that different isoforms have different kinetics forpen, inactivated, and closed states implies differentialensitivity to drugs exhibiting state-dependent binding,nd there is indeed evidence for diversity of local anes-hetic effects on the different isoforms. For example, theammalian cardiac channel, Nav1.5, appears to be muchore sensitive than the neuronal channels, eg, Nav1.2, to

idocaine block. This difference in potency results from aifference in the levels of resting activation between thechannels and therefore reflects gating differences

ather than drug affinity differences between the chan-els.161,213,221 In vitro experiments using the Nav1.7 andav1.8 isoforms coexpressed with the �1 subunit show

hat Nav1.8 is 4 times more sensitive to lidocaine.46

Moreover, the isoforms present different potentialubstrates for several protein kinases, and some may alsoe directly modified by G-proteins.35 Exposure of neu-ons to inflammatory mediators (eg, prostaglandin (PG)

2), algogenic substances (eg, endothelin-1), and neuro-ransmitters (eg, serotonin) can modify TTX-R channelating through enzymatic phosphorylation of intracellu-

ar portions of the �- and �-subunits. However, there isittle evidence that TTX-S channels are subject to such

igure 3. Results of a computer simulation of the responses ofneuron that expresses only sodium channels with fast inacti-

ation rates (A) and the change produced by switching 5% ofhe channels to the slowly inactivating type (B). Responses are to

single .1-msec depolarizing stimulus. Data from Gent et al,
npublished observations.

rasitp

tlnvtWipns

TC

saegntwwuacwa

P

d

wlntthfatpttwcj(

twTm

ttaswtfastn

iff

FabD*mh


eceptor-driven phosphorylation in primary nociceptors,lthough such channels and their related currents areensitive to direct activation of protein kinase (PK) A andts intracellular pathways in cultured cells.35 The extento which these mechanisms contribute to neuropathicain in vivo is unknown.The differential distribution of channel isoforms be-

ween different tissues raises the hope that isoform-se-ective drugs that modify function in primary afferenteurons may be relatively free of the toxic central ner-ous system (CNS) and cardiac side effects that constrainhe use of currently available sodium-channel blockers.

e focus here on the role of neuronal sodium-channelsoforms, but we note that sodium channels are also ex-ressed by glia cells in both the peripheral and centralervous systems; glia cells may also be a target ofodium-channel blocking drugs.192

ricyclic Antidepressants as Sodium-hannel BlockersAny discussion of the role of sodium channels in pain

tates involves evidence derived from the use of localnesthetic drugs. However, we must also consider thevidence from tricyclic antidepressants (TCAs). Thisroup of drugs is commonly used in the treatment ofeuropathic and other chronic pain states.32,153 All ofhe TCAs are known to have a complex mixture of actionshich might contribute to their analgesic action. Earlyork concentrated on their ability to inhibit the re-ptake of norepinephrine and serotonin.153 TCAs havelso been shown to block sodium, potassium, and cal-ium voltage-gated channels.167 We focus here onhether activity at sodium channels contributes to theirnalgesic efficacy.

otent Blockers of Sodium ChannelsCompared to bupivacaine, many TCAs show a longer

igure 4. (A) Time of complete proprioceptive, motor, and nopplied to rat sciatic nerve. �: Block durations of all 3 functioupivacaine group. *: Nociceptive blockade is significantly longeata from Gerner et al.85 (B) Long-lasting analgesia in human (placebo vs 100 mmol/L amitriptyline), ** (10 vs 50 mmol/L amitmol/L amitriptyline). Reprinted with permission from Gernerealthy volunteers. Reg Anesth Pain Med 28:289-293, 2003.84

uration of blockade of nerve impulse propagation e

hen tested in animal preparations (Fig 4). Amitripty-ine was more potent than bupivacaine in a subcuta-eous infiltration model126 and when applied intra-hecally in rat and sheep.83 However, when ami-riptyline was evaluated for ulnar nerve blockade inealthy human volunteers, it was found to be less ef-ective than bupivacaine— contrary to the results fromlarge number of animal studies. This might be due to

he thicker nerve sheaths present in humans as com-ared to rats, presenting a larger barrier to penetra-ion for amitriptyline into the nerve.80 When appliedopically, the pain-blocking activity of amitriptylineas found to be significantly more effective than pla-

ebo in healthy human volunteers,84 and some sub-ects had a complete analgesia lasting several hours Fig 4).Local anesthetics and TCAs both bind more tightly to

he inactivated state of the sodium channel.34 Thus, asith local anesthetics the nerve block obtained withCAs is use dependent, and this dependency may be evenore pronounced for TCAs.167

There are data suggesting that various TCAs differ inheir ability to block sodium channels. Whereas amitrip-yline, doxepin, and imipramine were superior to bupiv-caine in blocking nerve impulse propagation in a ratciatic nerve preparation, trimipramine and desipramineere somewhat less effective than bupivacaine, and nor-

riptyline, protriptyline, and maprotiline were clearly in-erior.198 However, it is unclear whether the differencesmong TCAs are due to differences in their activity atodium channels or to differences in their ability to passhrough various membrane barriers within peripheralerve trunks.It remains to be seen whether various TCAs show sim-

lar binding affinities for the various sodium-channel iso-orms. There is evidence for 2 distinct binding affinitiesor bupivacaine and amitriptyline at physiologically rel-

tive blockade produced by bupivacaine vs amitriptyline whenthe amitriptyline groups are significantly longer than in the

n motor blockade in the 5 mmol/L amitriptyline group (P � .05).nteers with a topical preparation of amitriptyline. P � .05 forline), *** (10 vs 100 mmol/L amitriptyline), and � (placebo vs 50o G, Srinivasa V, Narang S, Wang GK: Topical amitriptyline in

cicepns inr thavoluriptyP, Ka

vant membrane potentials.85

CN

C

pdnongniraadnclpTs

E

anm1aDpdordnfpceeocrctatirnoalDto

CfNb

E

mmnmiDidoaspm

vadmiottmo

Fictml(rtqafec


ontribution of Sodium Channels toeuropathic Pain Conditions

linical EvidenceSeveral clinical studies have demonstrated that thera-eutic agents that exhibit use-dependent block of so-ium channels show efficacy against painful peripheraleuropathy. Systemic administration of lidocaine andther sodium-channel blockers relieves the symptoms ofeuropathic pain in patients with postherpetic neural-ia, painful diabetic neuropathy, idiopathic trigeminaleuralgia, and other conditions.12,74,123,149,184,189 A top-

cal preparation of lidocaine relieves postherpetic neu-algia,182,183 and preliminary studies suggest that it maylso have efficacy in painful diabetic peripheral neurop-thy.11 Intravenous lidocaine has shown efficacy in aouble-blind placebo-controlled study in patients witheuropathic pain following spinal cord injury.75 Sodiumhannel blockade is a likely mechanism through which ateast some antiepileptics might suppress neuropathicain. As noted above, the well established efficacy ofCAs may be due, at least in part, to their ability to blockodium channels.

xperimental EvidenceThe expression of sodium-channel isoforms in primary

fferent neurons is dramatically altered after peripheralerve transection and in animal models of post-trau-atic and diabetic painful peripheral neuropathy (Table

). Nav1.1 and Nav1.2 are down-regulated followingxotomy.130 Nav1.3 is not normally expressed in adultRG, but its expression is greatly up-regulated after com-lete and partial nerve injuries and after the induction ofiabetes.22,51,63,130,167,191,217 In contrast, dorsal rhizot-my, which severs the central branch of the DRG neu-on’s axon, has no effect on Nav1.3 expression.22,217 Thee novo expression seen after nerve transection can beormalized by treatment with glia-derived neurotrophicactor (GDNF) and nerve growth factor (NGF).139 The ex-ression of Nav1.6 is up-regulated in small and large DRGells in experimental diabetic painful neuropathy.51 Thexpression of Nav1.8 is down-regulated in primary affer-nt neurons following transection of their peripheral ax-ns, but no change is seen when their central axons areut.191 In partial nerve injuries, the intact afferent neu-ons (perhaps only the C-nociceptors) show little or nohange in the expression of Nav1.890 but redistributehese channels from their cell bodies in the DRG to theirxons.90,160 Contrary results have appeared concerninghe fate of Nav1.8 expression in the chronic constrictionnjury (CCI) model:17 A decrease in small cells has beeneported by Dib-Hajj et al,63 but Novakovic et al160 foundo change. In the CCI model, the DRG contain a mixturef axotomized and intact neurons; it is likely that there isdecrease of Nav1.8, but only in the axotomized popu-

ation.57 Nav1.8 is down-regulated in small and largeRG neurons in diabetic rats.51 Peripheral nerve transec-

ion, but not dorsal rhizotomy, decreases the expression
f Nav1.9.191 A decrease is also seen in small cells in the e
CI model63; but in this case the change is also probablyound only in the axotomized cells.57 Interestingly,av1.9 is increased in both small and large cells in dia-etic rats.51

ctopic DischargeA large body of evidence suggests that the develop-ent of discharge from ectopic foci after nerve injuryay represent an underlying mechanism that drives

europathic pain.1,60,98 These abnormal dischargesay arise at the site of nerve injury or in the

njured primary afferent neuron’s cell body in theRG.58,119,120,132,151,210,211 Moreover, it is now clear that

n partial nerve injuries, intact axons also have ectopicischarge when they are neighbors to degenerating ax-ns.222 Electrophysiologic studies show that both myelin-ted and unmyelinated primary afferent axons showpontaneous activity after nerve injury, and it is thusrobable that such discharge is present in low-thresholdechanoreceptors and in nociceptors.60

Seminal discoveries in this field were reported by De-or et al.59,61 First, they showed that sodium channelsccumulate at the stump of severed axons. Second, theyemonstrated that ectopic discharge in axotomized pri-ary afferent neurons is very sensitive to lidocaine, be-

ng blocked by doses that have no effect on the initiationf nerve impulses by natural stimuli or on the propaga-ion of nerve impulses. Moreover, they showed that ec-opic discharge originating in the DRG is about 5 timesore sensitive to lidocaine than the ectopic discharge

riginating at the site of nerve injury (Fig 5). There is

igure 5. Cumulative dose of lidocaine (repeated IV bolus dos-ng) producing a complete block of spontaneous ectopic dis-harge in primary afferent neurons recorded after sciatic nerveransaction in the rat. Ectopic discharge arising from the axoto-ized neuron’s cell body in the DRG is 4-5 more sensitive to

idocaine than discharge arising from the site of nerve injuryneuroma). Note that the graph shows the percentage of neu-ons whose spontaneous discharge was completely blocked byhe indicated dose. A significant reduction in discharge fre-uency would be obtained by plasma levels less than thosechieved with the indicated dose. Reprinted with permissionrom Devor M, Wall PD, Catalan N: Systemic lidocaine silencesctopic neuroma and DRG discharge without blocking nerveonduction. Pain 48:261-268, 1992.61

vidence that mexiletine is even more potent than lido-

cfo

pasptonpncfarast

tmoclmeaadppipt

C

N

artpamsmraTtsNe

Nae

ac

N

grcCnmdtal

seiaondpeckcincf

N

tmssiwrlzptoar

oco

tjttb


aine in blocking ectopic discharge.40 These effects wereound at doses that are far below those that yield cardio-r neurotoxicity.The presence of spontaneous ectopic discharges mayrovide a mechanism to account for spontaneous painnd dysesthesiae after nerve injury. However, there areeveral caveats associated with this hypothesis. First, theresence of spontaneous discharge in injured fibers, or inheir uninjured neighbors, may play a role in spontane-us pain. But work with experimental animal models ofeuropathic pain rarely measures spontaneous pain. Ex-erimentally, we usually measure the response of aerve-injured animal with light touch, noxious heat, orold stimuli, and changes in the threshold for the noci-ensive reflex response to these stimuli are indices ofllodynia or hyperalgesia. Many, but perhaps not all, pe-ipheral neuropathy patients have spontaneous pain inddition to various kinds of allodynia and hyperalge-ia,16 but we do not know if there is a causal link be-ween spontaneous and stimulus-evoked pains.A second argument against the interpretation of spon-

aneous discharge of injured afferents being “the”echanism of spontaneous neuropathic pain is the issuef time dependency. It is known that spontaneous dis-harges occur soon after injury but, in the animals ateast, they diminish rapidly and appear to revert to nor-

al within 20 to 50 days, whereas behavioral signs ofvoked neuropathic pain are present for many weeksfterwards.19,103,145,148 Even during the period immedi-tely following nerve injury, the correlation between theegree of spontaneous activity and evoked neuropathicain behaviors may be weak.72,143-145 These findings em-hasize the distinction between the mechanisms which

nitiate neuropathic pain and those that maintain it. It isossible that the maintenance phase relies on the evolu-ion of CNS mechanisms.33,70,98,118

hannel Isoforms and Neuropathic Pain

av1.3At least some primary afferent ectopic discharge and

t least some of the symptoms of painful peripheral neu-opathy are TTX sensitive.164 For example, low concen-rations of TTX, which have no effect on nerve impulseropagation, block spontaneous ectopic dischargend mechanoallodynia in the spinal nerve ligationodel.47,146 Nav1.3 is TTX sensitive, and its expression in

mall, presumably nociceptive, DRG neurons (as well as inedium and large, presumably nonnociceptive, neu-

ons) is up-regulated following axotomy in the SNL, CCI,nd diabetic neuropathy models.22,47,51,63,115,130,141,217

he post-injury increase in Nav1.3 expression parallelshe appearance of a rapidly repriming sodium current inmall DRG neurons.22 These observations suggest thatav1.3 may make a key contribution to neuronal hyper-xcitability.Post–nerve injury treatment with GDNF normalizesav1.3 expression, reduces ectopic discharge in A-fibers,nd reduces pain.27,212 However, a 50% decrease in the
xpression of Nav1.3 in rat DRG cells, obtained via the a
ntisense method, has no effect on mechanoallodynia orold-allodynia in the rat spared-nerve injury model.141

av1.7Nassar et al157 have produced mice in which the Nav1.7ene is knocked out selectively in nociceptive DRG neu-ons. These animals do not differ from their wild-typeontrols in the development of mechanoallodynia in thehung model131 of post-traumatic painful peripheraleuropathy. Moreover, mechanoallodynia develops nor-ally in mice in which both Nav1.7 and Nav1.8 have been

eleted.156 However, expression of Nav1.7 is increased inhe DRG of rats with streptozotocin-induced painful di-betic neuropathy, and there is also an increase in theevel of phosphorylation of the channel.115

Erythromelalgia (also “erythermalgia”) is a rare auto-omal dominant condition in which the patient experi-nces episodes of burning pain and red hot skin, usuallyn the extremities and usually exacerbated by warmthnd relieved by cooling. This condition has been vari-usly ascribed to neuropathic and vasomotor mecha-isms and to small-fiber neuropathy.55,166 There is nowefinitive evidence that at least some erythromelalgiaatients have an inherited mutation in the gene thatncodes the Nav1.7 channel.64,69,155,231 The mutanthannel has alterations in its activation and deactivationinetics, a lowered threshold for activation, and an in-rease in current amplitude, resulting in hyperexcitabil-ty and high-frequency discharges in nociceptive DRGeurons.53,64 It remains to be seen whether similarhanges in Nav1.7 function are present in acquired pain-ul peripheral neuropathy.

av1.8Several lines of evidence suggest that Nav1.8 is impor-

ant in painful peripheral neuropathy. Nav1.8 makes aajor contribution to the TTX-resistant current found in

mall, presumably nociceptive, DRG neurons. Kral et al135

howed that the TTX-resistant current in small DRG cellssolated from CCI rats had activation thresholds thatere more negative than normal and that there was a

eduction in the average density of this current. DRGevels of Nav1.8 are down-regulated in rats with strepto-otocin-induced diabetes at times when the animals dis-lay mechanoallodynia.51,115 In the rat spinal nerve liga-ion (SNL) model of Kim and Chung,131 where the axonsf L5 and L6 DRG neurons are transected, Nav1.8 levelsre decreased in L5 DRG, but increased in the intact neu-ons of L4.234

Treatment with antisense, but not mismatch, oligode-xynucleotide (ODN) to Nav1.8 significantly reducedhannel expression and produced a temporary reversalf allodynia and hyperalgesia in SNL rats (Fig 6).90,136

Because the expression of Nav1.8 is down-regulated inhe L5/L6 small cell bodies whose axons have been in-ured in this model,191 the site of action of Nav1.8 inhibi-ion may lie elsewhere.90,136 Following axotomy and inhe CCI model, stores of Nav1.8 within the neuron’s cellody in the DRG are translocated to the neuron’s axon,
nd this correlates with an increase in TTX resistance in

stueisna

trnrmobcim

cttstnofit1N

ifnrHta

N

atifnwwotp

ME

p(tpdlrqt

TS

nnhbp(7u

FnndltbmMhaKnr


lowly conducting axons (almost certainly nocicep-ors).90,160 This translocation has also been seen in theninjured L4 DRG neurons in the SNL model, and there isvidence of ectopic discharge in the axons of L4 neuronsn that model. Thus, the pain-relieving effect of anti-ense-mediated inhibition of Nav1.8 in models of partialerve injury is likely due to changes in the uninjuredfferent axons.In contrast to these results, other evidence indicates

hat Nav1.8 is not essential for the development of neu-opathic pain. Kerr et al125 found no difference betweenormal mice and Nav1.8-null mutants, and Nassar et al156

eported no effect on mechanoallodynia in the Chungodel in Nav1.8 null mice or in mice with a double knock-

ut for Nav1.8 and Nav1.7. The reason for the discrepancyetween the antisense and knockout experiments is un-lear. It may be of importance that the antisense exper-ments were done in rats, whereas the knockout experi-

ents were done in mice.

igure 6. Antisense targeting of Nav1.8. Rats with the spinalerve ligation (SNL) model of post-traumatic painful peripheraleuropathy developed the expected ipsilateral mechanoallo-ynia (Top) and heat-hyperalgesia (Bottom), as shown at base-

ine (BL) testing. One day later, the animals were injected intra-hecally with antisense (AS) oligodeoxynucleotide (ODN) tolock the expression of the Nav1.8 sodium channel, with mis-atch (MM) ODN as a control or saline as an additional control.M and saline had no effect, but AS reversed the allodynia and

yperalgesia on the nerve-injured side for as long as it wasdministered. Reprinted with permission from Lai J, Gold MS,im CS, Bian D, Ossipov MH, Hunter JC, Porreca F: Inhibitions ofeuropathic pain by decreased expression of the tetrodotoxin-esistant sodium channel, NaV1.8. Pain 95:143-152, 2002.136

The possibility that the Nav1.8 is critical for ectopic dis- a

harge in injured primary afferent fibers is supported byhe results of experiments in genetically modified micehat lack the Nav1.8 isoform. In the wild-type mousetrain from which the mutants were prepared, section ofhe saphenous nerve and the subsequent formation of aerve-stump neuroma is associated with the appearancef a high incidence of ectopic discharge in A- and C-bers. By 3 weeks after nerve injury, 20% of the fibers inhe wild-type mice had ectopic discharge, but less than% of the fibers had ectopic discharge in mice lackingav1.8.185

It is important to note that there is evidence suggest-ng that the function of Nav1.8 may change with timeollowing nerve injury. In the first 2-3 weeks followingerve injury, the channel appears to contribute to neu-opathic pain by enabling activity in uninjured afferents.owever, with time, expression of the channel appears

o be up-regulated again in injured axons and may evenccumulate at the site of nerve injury.50

av1.9The persistent current carried by the Nav1.9 channel is

bundantly expressed in small DRG neurons, particularlyhe nonpeptidergic IB4-binding neurons that terminaten the spinal cord substantia gelatinosa. Mice lackingunctional Nav1.9 channels have no apparent change inormal pain sensation. Moreover, tests in mutant miceith nonfunctional Nav1.9 channels174 and in micehere the channel’s expression has been inhibited by theligodeoxynucleotide antisense method172 have failedo show any effect on the expression of neuropathicain.

embrane Potential Oscillations andctopic DischargeOne might hypothesize that the ectopic discharge inrimary afferent neurons results from the classicHodgkin-Huxley) repetitive firing process in which a sus-ained depolarization repeatedly draws the membraneotential above threshold. However, accumulating evi-ence acquired both in vitro and in vivo suggests that at

east in the cell bodies of primary afferent sensory neu-ons in the DRG, the ectopic discharge results from auite different process—subthreshold membrane poten-ial oscillations.6-8,122,143,168,223,226,227

he Oscillatory Mechanism Is VoltageensitiveWhen first penetrated by a microelectrode, most DRGeurons have a stable resting potential. However, a mi-ority (about 5%-10%) of the A-neurons (those thatave myelinated axons) exhibit periodic sinusoidal mem-rane potential oscillations of small amplitude (�2 mVeak-to-peak) with a mean frequency of about 100 HzFig 7).7,143,227 The oscillations are voltage sensitive (Fig): Both their prevalence and amplitude are enhancedpon depolarization.7,143,168 Accordingly, the percent-
ge of oscillating A-neurons is increased from about 6%

atav(r

NM

Dplptf

TfD

tcv

nepscclttsCottlhfatphcfi

FCCrf tionsa t ©1


t the resting potential to about 13% upon depolariza-ion (pooled data from experiments with both immaturend adult rats).7,143,144 In addition, oscillations are re-ealed in about 27% of the depolarized C-neuronsthose with unmyelinated axons) (immature and adultats).7

erve Injury Enhances the Oscillatoryechanism

Nerve injury induces an increase in the prevalence ofRG neurons with subthreshold oscillations. For exam-le, the proportion of depolarized A-neurons with oscil-

ations is increased from about 10% to 24% (adult rats,ooled data from both sciatic and spinal nerve transec-ion),7,143 and in C-neurons the proportion increasedrom 28% to 44% (adult rats, sciatic transection).7

he Oscillatory Mechanism Is Essentialor the Generation of the EctopicischargeThe oscillatory mechanism is essential to the genera-

ion of repetitive ectopic discharge.7,143,144,227 Only os-illating neurons show repetitive ectopic firing; con-

igure 7. Subthreshold sinusoidal membrane potential oscillat-neurons at various membrane potentials. Vr, resting potent-neuron. (Right) Percentage of DRG A- and C-neurons with meesting potential (open bars) and when depolarized (solid bars)rom Amir R, Michaelis M, Devor M: Membrane potential oscilland neuropathic pain. J Neurosci 19:8589-8596, 1999.7 Copyrigh

ersely, neurons without subthreshold oscillations do (

ot, even upon depolarization.7,143 Some neurons firectopic singlet spikes at low frequency with an irregularattern. Each such spike is triggered by the upstroke of ainusoidal oscillation that crosses threshold, indicating aausal relation between the oscillations and the dis-harge.7,143 Other neurons fire sustained trains of regu-arly spaced spikes, or sequences of spike bursts. Here,he first impulse in the train (or burst) is consistentlyriggered by an oscillation sinusoid, and each subsequentpike is triggered by a depolarizing after-potential.8

onsistent with the hypothesis, because the prevalencef oscillating neurons is relatively low at resting poten-ial, spike discharge is also rare, with only about 4% ofhe A-neurons firing spontaneously. However, on depo-arization, as the underlying oscillatory mechanism is en-anced, this percentage is increased to 11% (pooled datarom both immature and adult rats).7,143 In addition,bout 8% of the C-cells begin to fire repetitively (imma-ure and adult rats).7 Because nerve injury increases therevalence of oscillating neurons, it consequently en-ances the prevalence of ectopically active neurons. Ac-ordingly, the proportion of depolarized A-neurons thatre repetitively is increased from about 10% to 23%

from noninjured rat DRG neurons. (Left) Oscillations in A- andote ectopic impulses arising from 2 of the oscillations in thene potential oscillations and ectopic discharge at the neuron’s.001 vs control. Reprinted with permission (left side of figure)

in dorsal root ganglion neurons: Role in normal electrogenesis999 by the Society for Neuroscience.

ionsial. Nmbra. *P �

adult rats, pooled data from both sciatic and spinal

nit

cptsTs(atfitwt

SGP

nTtCsscrpntspacewltepb

hmdTFdcfTatmAi

sarotp

StM

mhctTmpiaeNowgfi

HNN

cbdcrniadtwpdbcelHrNidqstcd


erve transection),7,143 and in C-neurons the proportionncreases from 12% to 27% (adult rats, sciatic transec-ion).7

Thus, the question of what causes repetitive ectopic dis-harge becomes 2 questions: 1) What causes membraneotential oscillations? and 2) what causes the depolariza-ions that raise them to the firing threshold? There areeveral processes capable of depolarizing DRG neurons.hese include mechanical stress (eg, during movement ortraight-leg raising),159 cell-to-cell cross-depolarizationwhere activity in a subpopulation of DRG neurons affectsdjacent neurons),4,5,8 and sympathetic efferent activi-y.8,171,226 Indeed, all of these mechanisms increase ectopicring in vivo.60 Oscillatory behavior in DRG neurons can

hus be thought of as a “motor” ready to be engagedhen the “clutch” of a slow-onset physiologic depolariza-

ion is released.

odium Channels Are Important for theeneration of Subthreshold Membraneotential OscillationsThe ionic basis of the oscillatory mechanism in DRGeurons has been studied.6,7 A voltage-sensitive andTX-sensitive sodium conductance appears to be impor-ant for the rising phase of the oscillations in both A- and-neurons. The oscillations are eliminated by partial sub-titution of sodium ions with choline in the perfusionolution, and by bath application of TTX or lido-aine.7,168,227 Once eliminated, oscillations could not beestored by depolarization, but they generally reap-eared following washout of the blocker. Calcium-chan-el blockers were not effective at blocking the oscilla-ions.7,168 Suppression of oscillations with sodiumubstitution, or via TTX or lidocaine block, also sup-ressed ectopic firing, consistent with the idea of a caus-tive relation between the oscillations, the ectopic dis-harge, and sodium channels. Oscillations and resultingctopic discharge were consistently blocked at timeshen propagation of impulses evoked by axonal stimu-

ation persisted, indicating that the effect was specific tohe ectopic discharge mechanism, rather than a local an-sthetic effect.7,227 Ectopic firing in vivo is also sup-ressed using lidocaine concentrations insufficient tolock axon conduction.61

Although there are considerable data supporting theypothesis that a TTX-sensitive channel underlies theembrane potential oscillations, there are 2 pieces ofata to suggest that, at least under certain conditions,TX-resistant current may also contribute to oscillations.irst, depolarization-potentiated oscillations have beenescribed in trigeminal ganglion neurons, and these os-illations are resistant to TTX. In contrast, the oscillationsrom DRG neurons are persistent and TTX sensitive.176

his may correlate with a difference between trigeminalnd DRG neurons; in the rat, trigeminal nerve transec-ion evokes relatively little ectopic discharge.202 Second,embrane potential oscillations induced in dissociated-neurons appear to have a voltage dependence that is

nconsistent with the biophysical properties of TTX-sen- D

itive channels present in sensory neurons.142 The volt-ge dependence of oscillation amplitude in these neu-ons peaks at about �20 mV, a potential at which 100%f TTX-sensitive channels would be inactivated andherefore unable to contribute to the depolarizationhase of the oscillation.

odium Channel Isoforms That Are Likelyo Contribute to the Oscillatoryechanism

Among the sodium channel subtypes, Nav1.3 is theost likely to contribute to the oscillatory mechanism. It

as fast enough gating kinetics, it is up-regulated at pre-isely the same time after axotomy that oscillations, ec-opic firing, and behavioral allodynia commence, andTX suppresses all 4.22,144,217 Of course, this does notean that other sodium-channel isoforms might not

lay important roles. Selective block of any one of thesoforms might be enough to attenuate the oscillationsnd bring at least some cells below the threshold forctopic firing. The TTX-resistant isoforms Nav1.8 andav1.9 are particularly interesting in this regard. More-ver, the relatively persistent depolarizations producedhen Nav1.8 and Nav1.9 open may be important both inenerating the oscillations and in bringing them to thering threshold.

yperpolarization-Activated, Cation-onselective, Cyclicucleotide–Modulated ChannelsIt is known that sodium currents in DRG neurons are

arried not only by the channels that respond to mem-rane depolarization but also by a family of channelsesignated as HCN—for hyperpolarization-activated,ation-nonselective, cyclic nucleotide–modulated—thatespond to membrane hyperpolarization. These chan-els pass both sodium and potassium. They are involved

n cardiac pacemaker activity, and in DRG neurons theyre known to be involved in the production of rhythmicischarge and to modulate the membrane’s resting po-ential.68,207 Four isoforms have been identified, ofhich HCN1-3 are clearly expressed in rat DRG cells; theresence of HCN4 is uncertain.42,207 Large- and medium-iameter DRG neurons that probably correspond to fi-ers with A� and A� conduction velocities express thesehannels abundantly. However, about 15% of DRG cellsxpressing HCN also express TRPV1, suggesting that ateast some C-nociceptors may also express this channel.68

CN1 and HCN2 are significantly, but partially, down-egulated in the L5/L6 DRG of rats in the SNL model.evertheless, in a rat nerve-ligation model, ZD7288, an

nhibitor that is specific for HCN channels, significantlyecreases mechanoallodynia and the ectopic firing fre-uency in both A� and A� fibers (by 90% and 40%, re-pectively).42 There is also evidence that HCN channels inhe hippocampus have a role in epilepsy.14 The pharma-ologic characteristics of HCN channels are not known inetail. Local anesthetics block HCN channels in rat small
RG neurons at clinically relevant concentrations.21 The

dpma

ccvvacorrprm

S

CNidreadOoccrNicma

dpwcdplbhwt

tssnttdh

criaahsnCtctwntcctt

nsdrndhntmimefiftuacn

PD

jtcmgcittpwwaa


ata from SNL rats and the clear role that HCN channelslay in rhythmic repetitive firing and in regulating theembrane resting potential suggest that they may haverole in neuropathic pain.However, it does not seem likely that HCN channels

ontribute to the subthreshold membrane potential os-illations described above. First, the current conductedia HCN channels in DRG neurons is minimal or inacti-ated at membrane potentials required for oscillationctivity. Second, the current carried by HCN channels isonsistently inhibited upon depolarization, whereas thescillatory mechanism is enhanced. Third, the HCN cur-ent has very slow kinetics, which is inconsistent with theelatively high frequency of the subthreshold membraneotential oscillations. Fourth, HCN1 and HCN2 are down-egulated following SNL at times when the oscillatoryechanism is enhanced.42,143

odium Channels on CNS NeuronsThere is evidence that sodium channels expressed byNS neurons also may be involved in neuropathic pain.av1.3 is up-regulated in spinal cord dorsal horn neurons

n CCI rats. Inhibiting this increase with antisense ODNecreased dorsal horn neuron hyper-responsiveness andeduced the animal’s abnormal pain responses.101 Thexpression of Nav1.3 is increased in dorsal horn neuronslso following a traumatic spinal cord injury that pro-uced the central pain syndrome. Here too, the antisenseDN procedure decreased neuronal and behavioral signsf abnormal pain.100 Very recent work shows that spinalord injury causes an increase in the excitability of noci-eptive neurons in the ventrobasal thalamus and an up-egulation of Nav1.3 in thalamic neurons. Inhibition ofav1.3 expression by antisense ODN given via lumbar

ntrathecal administration reversed the excitabilityhanges and decreased Nav1.3 expression in the thala-us.102 It is not clear how treatment at the spinal level

ffected thalamic expression levels.Microinjection of lidocaine into the rostroventral me-ial medulla (RVM) blocks both tactile and thermal hy-eresthesia134,170 when tested 4 days and later but notithin the first 2 days after SNL. The time course of lido-

aine activity emphasizes the possibility of a time-depen-ent plasticity in the CNS that may maintain neuropathicain. The systemic delivery of quarternary derivatives of

idocaine, QX222, and QX314, which do not cross thelood-brain barrier, has been shown to reverse thermalyperalgesia,164 but not tactile hyperesthesia, in ratsith SNL.44 In contrast, QX314 reversed tactile hyperes-

hesia when microinjected into the RVM.44

The mechanisms by which the QX compounds achieveheir effects are not fully understood. Voltage-clamptudies in both myelinated and nonmyelinated axonshow that the QX compounds block TTX-S sodium chan-els only from the intracellular compartment,79b,194b yethey reduce the C-fiber elevation of the compound ac-ion potential of rat nerves164 with a strong “frequency-ependent” inhibition; frequency-dependent block is a
allmark of local anesthetic drug actions on sodium a
hannels.194b Systemically administered QX compoundselieve symptoms of neuropathic pain in rats and reducenjury-related ectopic impulse activity, with impulsesrising from the injury site and the DRG being consider-bly more sensitive than those recorded in the dorsalorn. A possible interpretation of these results is thatystemic QX compounds have easier access to peripheralerve sodium channels but are much slower to reach theNS. One might speculate that the sites of action forhese “therapeutic” effects are primarily TTX-R sodiumhannels, whose structure at the outer pore differs fromhat at TTX-S channels, similar to the differencesrought by site-directed mutagenesis of sodium chan-els that allows extracellular local anesthetics to reachheir blocking site from the outer opening of thehannel.176b However, whether the actions of the QXompounds truly involve only sodium channels and, if so,he degree of isoform selectivity of this effect are impor-ant questions that remain unanswered.The observations described suggest that sodium chan-els may have different roles in the different kinds oftimulus-evoked pain abnormalities (ie, mechanoallo-ynia evoked by input from A� low-threshold mechano-eceptors vs heat-hyperalgesia evoked by input from C-ociceptors). This is consistent with the idea that theifferent kinds of neuropathic stimulus-evoked painsave at least partly different pathophysiologic mecha-isms. There is considerable evidence that this is true inhe animal models224 and some evidence that it is true inan. For example, using selective nerve block techniques

n human volunteers who received topical application ofustard oil, it has been shown that mechanoallodynia is

voked by input from the large-diameter myelinated A�bers, whereas heat-hyperalgesia is evoked by inputrom unmyelinated C-nociceptors.132 The possibility thatactile allodynia is mediated by large fibers raises thenanswered question of how “knockdown” of Nav1.8 byntisense ODN can produce antiallodynic actions, be-ause Nav1.8 is almost exclusively found on small DRGeurons in normal animals.

eripheral Changes in Sodium Channelsrive CNS ChangesSeveral lines of evidence suggest that after nerve in-

ury, enhanced peripheral neuronal activity, includinghe ongoing discharge of C-nociceptors, results in centralhanges at spinal levels. Central sensitization219 is theost thoroughly studied example. Recent studies sug-

est that post-injury changes at supraspinal levels alsoontribute. For example, there is evidence that nervenjury is followed by activation of descending pain-facili-atory mechanisms in the brain stem that serve to main-ain pain hypersensitivity.173 As noted, Nav1.8 is foundrimarily in small-diameter nociceptors, yet knockdownith antisense ODN, or normalization of its distributionith growth factors (see below), reverses both mechano-llodynia and heat-hyperalgesia after SNL. Althoughcute systemic administration of QX314 was not active
gainst mechanoallodynia in nerve-injured rats, mech-

aatbtQ

vrthrtlAeicct

cwnctttrbtdoatc

LDB

oolwiswtdwtr

t�rgr

pclprhoafTswpprsclrol

nalatccdlfie

atia

pmlslaa

CI

ktrflwPi


noallodynia was reversed by long-term (several days)dministration of QX314.164 This suggests that long-erm suppression of peripheral activity with QX314 maye necessary to enable reversal of CNS changes, althoughhe result could also reflect a slow accumulation ofX314 in the CNS during systemic delivery.The first hypothesis is also supported by recent obser-

ations made with artemin, a member of the GDNF neu-otropic factor family.82 Prolonged systemic administra-ion of artemin reversed both mechanoallodynia andeat-hyperalgesia in rats with SNL.82 The injury-inducededistribution of Nav1.8 along the sciatic nerve was re-urned to the normal state by artemin, as was up-regu-ation of spinal dynorphin and other neuropeptides.82

rtemin acts through the GFR�3 receptor, which is foundxclusively on small diameter nociceptors.165 These find-ngs support the concept that chronic treatments thatan interfere with peripheral changes after nerve injuryan result in a cascade of events that serve to normalizehe chronic pain state.It is important to note that accumulating evidence indi-

ates that neuropathic pain is a chronic abnormal statehose mechanisms evolve over time. After peripheralerve injury, a number of central and peripheral neuralhanges occur that are essential for the initiation and main-enance of the neuropathic pain state. Either the preven-ion or the reversal of these central or peripheral adapta-ions may return the nervous system to a normalized state,esulting in prevention of pain expression. Sodium-channellockade, over an extended time period, might accomplishhis goal and might therefore be useful as a mechanism ofisease modification. Critically, although the primary sitef action of sodium-channel blockers may be in the primaryfferent fiber, or perhaps within the CNS, the ultimate ac-ion may result from normalization of function at bothentral and peripheral sites.

ong-Lasting Pain Relief With Short-uration Exposure to Sodium-ChannellockersNeuropathic pain patients often report days or weeksf pain relief after brief treatments with lidocaine andther sodium-channel blockers. It is now firmly estab-

ished that an analogous phenomenon occurs in ratsith experimental peripheral neuropathies who receive

nfusions of lidocaine. For example, Chaplan et al41

howed that an intravenous infusion of lidocaine, whichas maintained at constant plasma levels for 30 minutes

o several hours, eliminated the animal’s mechanoallo-ynia not only during the infusion but also for at least 3eeks afterwards. In addition, Mao et al150 have shown

hat a single bupivacaine block of the sciatic nerve in CCIats relieves heat-hyperalgesia for at least 2 days.Strichartz et al10,190,196 have shown that there is a

hreshold lidocaine plasma concentration of 2.5-5mol/L required for long-lasting (48 h post-infusion) painelief. They also found that there was a limit to the de-ree of long-lasting relief attainable in each individual
at; neither repeated infusions to the same effective s
lasma concentration nor graduated infusions to in-reasingly higher levels (which were still well below toxicevels), delivered every 48 h, could elevate the amount ofain relief above an individual’s “ceiling level.” Someats achieved complete and persistent relief, but mostad a more limited recovery. A minority, about a quarterf those tested, responded insubstantially to lidocainend had no relief during or after the infusion. The reasonor these individual differences has yet to be identified.here was a strong correlation between the acute rever-al of allodynia during the infusion and the degree tohich it was sustained 2 days later, implying that these 2henomena are mechanistically coupled. In identical ex-eriments with mexiletine, the same degree of acuteelief during the infusion was never followed by a per-istent recovery, suggesting a special activity for lido-aine. Because the half-life of lidocaine in the rat’s circu-ation is at best a few hours, and its less potent (withespect to impulse blockade) metabolites are present fornly a few hours longer,124 the enduring recovery is un-

ikely to be due to any residual drug.Detailed examination of the time course of this phe-omenon revealed 3 distinct phases of activity (Fig 8). Innimals infused 2 days after nerve injury, substantial re-ief was seen during the 30-min infusion (with mechano-llodynia reduced by about 70%). Thirty minutes afterhe end of the infusion, mechanoallodynia returnedompletely. Remarkably, beginning about 6 h later andontinuing over the next 24 h, mechanoallodynia slowlyisappeared again and remained at a greatly reduced

evel for 1 to 2 weeks. Therefore, the sustained recoveryrom allodynia was not a continuation of the relief dur-ng drug infusion, but was instead a delayed and slowlyvolving effect.Identical infusions of lidocaine at 7, rather than 2, days

fter injury gave essentially identical acute effects andhe same delayed rise in threshold over the next 24 h, butn this case the recovery was not sustained and the ther-peutic effect faded over the following week.In summary, these observations suggest that the lateersistent phase of pain relief obtained with lidocaine isechanistically distinct from that seen acutely, that the

ong-lasting pain relief may or may not be coupled toodium-channel blockade, and that the long-lasting re-ief may depend critically on the pathophysiologic mech-nisms underlying mechanoallodynia, with these mech-nisms varying over time and from animal to animal.

ontribution of Sodium Channels tonflammatory Pain ConditionsRecent evidence suggests that sodium channels play a

ey role in the pain and hypersensitivity associated withissue inflammation, and there is a possibility that thisole is especially important in the context of chronic in-ammation. A�- and C-nociceptors become sensitizedhen the tissue that they innervate becomes inflamed.rimary afferent nociceptor sensitization is character-zed by the appearance of 3 abnormal properties: 1)
pontaneous discharge, 2) a lowered threshold to the

attpppppnTttmmctbsnsmdl

C

tomm

fbcsep

E

obtimeoa

sflsvccpfw(t

FmsS hanici


dequate stimulus, and 3) a stimulus-response (S-R) func-ion that is shifted leftwards. These changes contributeo the generation of ongoing (“spontaneous”) pain, hy-eralgesia, and allodynia. Of course, ongoing pain, hy-eralgesia, and allodynia are also symptoms of neuro-athic pain. The question is therefore to what extent thehenomena found in both inflammatory and neuro-athic pain states share common physiologic mecha-isms, including contributions from sodium channels.issue inflammation also activates a class of nociceptorshat are not normally excitable. These “silent” nocicep-ors become excitable when exposed to an inflammatoryilieu; when activated, they behave like sensitized nor-al nociceptors.154 It is not known whether “silent” no-

iceptors are engaged in neuropathic pain states, andhere are no data on their sensitivity to sodium-channellockade. However, it is of interest to note that 90% ofilent C-nociceptors express Nav1.9.73 It is important toote that most of our knowledge about nociceptor sen-itization comes from experiments using acute inflam-atory states. As we note in the following, there is evi-ence that acute and chronic inflammatory pain have at

east partly different mechanisms.

linical EvidenceSodium-channel blockers are not usually considered in

he therapy for chronic inflammatory pain, but recentpen-label trials11,54,81 suggest that topical lidocaineay relieve low back pain, the pain of osteoarthritis, and

igure 8. Brief infusions of lidocaine (4 �g/mL plasma 14.7echanical allodynia ipsilateral to the ligation in 3 phases: 1,

ustained. Solid circles lidocaine infusion; open circles salintrichartz GR: Multiple phases of relief from experimental mecnfusions. Pain 103:21-29, 2003.10

yofascial pain. The mechanism that produces such ef- a

ects is unclear; it seems very unlikely that anestheticlock of cutaneous nociceptors could relieve pain that isoming from deep tissues. Topical lidocaine in thesetudies results in low plasma levels (�.2 �g/mL), so anffect can not be ascribed to blockade of nerve impulseropagation.

xperimental EvidenceThe inflammatory pain state created by a subcutane-us injection of complete Freund’s adjuvant (CFA) haseen studied extensively. Studies using an antiserumhat recognizes an epitope common to all Nav-channelsoforms have shown that CFA injection evokes a dra-

atic up-regulation of sodium channel expression. Theffect is of rapid onset, with peak levels of up-regulationccurring within about a day, and the effect persists fort least 2 months.94,97

There have been 2 studies of lidocaine’s effects onpontaneous discharge in nociceptors innervating in-amed tissue where the plasma levels have been mea-ured. Puig and Sorkin175 (Fig 9) used a continuous intra-enous infusion to examine lidocaine’s effects on 14utaneous C-fiber nociceptors with spontaneous dis-harge following formalin injection in the hind paw (ap-roximately 30 min before lidocaine). Decreases in firingrequency and complete block of spontaneous dischargeere seen with lidocaine plasma levels of 15-34 �mol/L

mean 21 �mol/L). The mean threshold lidocaine concen-ration was not reported, but effects were clearly seen at

ol/L) at 2 days after sciatic nerve ligation (SNL) in rat reverse-infusion; 2, transient postinfusion; and 3, slowly developingusion. Reprinted with permission from Araujo MC, Sinnott CJ,al allodynia by systemic lidocaine: Responses to early and late

�mperie inf

bout 6.5 �mol/L. Mechanical stimulation of the C-fiber’s

ct

eoaTppaadmpapwbmBloi�cn

srbet

twhp

C

N

tssrinataaodItaNt

N

um

FIlIw h pei pha


utaneous receptive field evoked discharge even whenhe ongoing discharge was completely inhibited.Xiao et al225 used a continuous intravenous infusion to

xamine lidocaine’s effects on 9 A�- and C-fiber cutane-us nociceptors with spontaneous discharge followingdministration of an antibody against GD2 ganglioside.his antibody is used to treat pediatric neuroblastomaatients; its infusion frequently causes the rapid onset ofain and mechanoallodynia. Intravenous infusion in ratslso causes mechanoallodynia of rapid onset. The mech-nism by which the antibody produces pain and allo-ynia is not known, but it is likely to involve a neuroim-une response with similarity to the neuroimmunehenomena that occur during inflammation. Xiao etl225 found that plasma lidocaine levels of 1.4-8.6 �mol/Lroduced a �50% decrease in spontaneous dischargeithout affecting the fiber’s mechanical threshold. Thelock of spontaneous discharge often persisted for manyinutes after the infusion (45-60 min in 2 extreme cases).ased on these 2 studies, the threshold plasma lidocaine

evel needed to block the ongoing discharge of cutane-us A�- and C-nociceptors in the context of an acute

nflammation is likely to be in the range of 1.3-6.5mol/L. Note that these were studies of cutaneous noci-eptors; there are no data on the lidocaine sensitivity ofociceptors innervating inflamed deep tissues.Behavioral indices of spontaneous pain and hyperalge-

ia in rats with an experimental arthritis are reduced byepeated systemic administration of the sodium-channellockers mexiletine and crobenetine.137 The analgesicffects became larger with repeated doses, suggesting

igure 9. Effect of lidocaine infusion on the ongoing inflammanfusion began 27 min after injecting formalin into the hind pawevels were estimated from the Stanpump program and confirmenhibition is first seen clearly at a concentration of about 1.5 �hen the spontaneous discharge was blocked. Reprinted wit

dentified primary afferent fibers: Systemic lidocaine suppresses

hat the mechanisms underlying the drug effect and/or s

he mechanisms of the pain and hyperalgesia evolvedith time. Systemically tolerable doses of mexiletineave also been shown to block behavioral and electro-hysiologic indices of pain in the formalin model.43,117

hannel Isoforms and Inflammatory Pain

av1.7 ChannelAn increase in the expression of Nav1.7 and its redistribu-

ion may contribute to inflammation-induced nociceptorensitization. Early experiments showed that a brief expo-ure of PC12 cells to NGF, which is known to play a majorole in inflammatory hyperalgesia,220 results in an increasen Nav1.7 expression and an increase in channel density ateurite terminals.205,206 Inflammation induced by CFA isccompanied by a rapid increase of Nav1.7 expression, andhis increase is blocked by pretreatment with cyclooxygen-se inhibitors.95 An increase of Nav1.7 expression is seenfter the administration of NGF.96 Nerve growth factor notnly increases expression of Nav1.7, but also increases theensity of the channel in sensory nerve terminals in vitro. 206

nflammation of the tooth pulp is associated with a redis-ribution of Nav1.7 to the nodes of Ranvier in thinly myelin-ted axons.193 Recent studies have shown that ablatingav1.7 in nociceptive neurons greatly reduces inflamma-

ory pain responses.156,157

av1.8 ChannelThere is a large body of evidence suggesting that mod-lation of Nav1.8 contributes to the initiation of inflam-ation-induced nociceptor sensitization and hyperalge-

-evoked discharge of a C-fiber cutaneous nociceptor in the rat.continued for the period marked by the lower arrows. Plasma3 plasma samples drawn at the times marked with open circles.

. Note that a mechanical stimulus still initiated discharge evenrmission from Puig S, Sorkin LS: Formalin-evoked activity inse-2 activity. Pain 64:345-355, 1995.175

tionand

d byg/mL

ia. Inflammation evoked by CFA is associated with a

ributtdcAdcntPcatwcpacmvtiaMabm

owtcpialNmdtg

eatcrvfgatmcin

CNMfpoesCalo

towtmiadetttcoapim

NtsdttmNisset

N

cIitdcmkrNa


apid increase in Nav1.8 expression in DRG that is inhib-ted by ibuprofen.95 The persistent hyperalgesia evokedy repeated injections of PGE2 is also accompanied by anp-regulation of Nav1.8 in DRG.209 Inflammatory media-ors induce a rapid increase in maximal sodium conduc-ance, a small hyperpolarizing shift in the voltage depen-ence of activation, and an acceleration of TTX-resistanturrent activation, all of which enhance excitability.87,88

large number of inflammatory mediators have beenemonstrated to induce these same changes in Nav1.8hannels: PGE2,71,88 serotonin,36,88 adenosine,88 epi-ephrine,128 endothelin-1,236 and NGF.234 In addition,he increased excitability of DGR neurons evoked byGE2 and forskolin in vitro is TTX resistant.71 The Nav1.8hannel might therefore act as a final common target for

diverse array of inflammatory mediators actinghrough a number of distinct second-messenger path-ays, including PKA,71,87 PKC,87,128 a nitric oxide,3 extra-

ellular signal-regulated kinase,66 and a ceramide-de-endent kinase.233,234 At least some of these effectsppear to reflect a direct phospholylation of the Nav1.8hannel.77 A role for Nav1.8 in the initiation of inflam-atory hyperalgesia is further supported by the obser-

ation that antisense knockdown of the channel blockshe development of hyperalgesia induced by a singlenjection of PGE2

127,209 as well as the long-lasting hyper-lgesia produced by repeated injections of PGE2.209

oreover, agonists of the �-subtype of opioid receptor86

nd group II metabotropic glutamate receptor230 blockoth the modulation of the channel and the develop-ent of inflammatory hyperalgesia140 and allodynia.229

Further evidence for a role for Nav1.8 in the early phasef inflammatory pain comes from studies in mice inhich its gene has been knocked out.2,138 For example,

hese animals have normal acute pain responses to vis-eral irritants (intracolonic instillation of saline and intra-eritoneal acetylcholine). However, intracolonic admin-

stration of the inflammatory and nociceptor-sensitizinggents capsaicin and mustard oil produced significantlyess pain-related behavior in the mutants that lackedav1.8 channels. Colonic inflammation in the wild-typeice was associated with referred hyperalgesia—a re-uction of pain thresholds to noxious stimuli applied tohe hind paw or abdominal skin. Mice lacking the Nav1.8ene had little or no referred hyperalgesia.Given that persistent inflammation is associated with

levated levels of inflammatory mediators, the associ-ted modulation of Nav1.8 is also likely to contribute tohe maintenance of inflammatory hyperalgesia. An in-rease in the expression of the channel may also have aole. Support for this suggestion comes from the obser-ation that there is an increase in Nav1.8 mRNA in DRGollowing peripheral injection of a high dose of carra-eenan and repeated injections of PGE2.24,203,209 There islso an increase in the relative density of the TTX-resis-ant current in visceral sensory neurons following inflam-ation of the gastric mucosa20 and colon.18 That these

hanges in expression may be associated with an increasen the transport of the channel to the peripheral termi-
als is supported by observations indicating that the m
FA-induced inflammation results in an increase inav1.8 in peripheral axons48 (but see Okuse et al162).oreover, a 4.5-fold increase in Nav1.8 levels has been

ound in the pulp extracted from painful teeth com-ared to pulp from nonpainful teeth180; the innervationf the dental pulp is almost entirely nociceptive. Furthervidence implicating Nav1.8 comes from experimentshowing that antisense knockdown of Nav1.8 attenuatesFA-induced hyperalgesia,172 the neuronal hyperexcit-bility observed in an animal model of cystitis,232 and the

ong-lasting hyperalgesia seen after repeated injectionsf PGE2 in rat.209

It should be noted that some of the data obtained inhe Nav 1.8 null mutant mouse2 do not support resultsbtained with other approaches. For example,hereas NGF-induced hyperalgesia is significantly at-

enuated, PGE2-induced hyperalgesia is unaffected inice lacking Nav 1.8.125 Moreover, the onset of CFA-

nduced hyperalgesia is delayed, but not inhibited,2

nd the hyperalgesia seen with cyclophosphamide-in-uced cystitis is unaffected.138 The basis for the differ-nt results obtained with mice lacking Nav1.8 andhose obtained with other experimental manipula-ions is not clear. It is possible that the genetic muta-ion that eliminates Nav1.8 induces compensatoryhanges in the expression of other ion channels orther mechanisms and that these compensatory mech-nisms mask the effects of the gene deletion.2 It is alsoossible that the involvement of Nav 1.8 in mediating

nflammatory pain varies with different inflammatoryediators.In fact, there is evidence to suggest that the role of theav1.8 channel differs during inflammation of different

issues. For example, both the proportion of afferentsensitized and the magnitude of the inflammation-in-uced sensitization of colonic sensory neurons197 appearo be greater than what is observed following inflamma-ion of cutaneous afferents.9 Both the magnitude of theodulation and the proportion of neurons in whichav1.8 is modulated are greater in colonic afferents than

n cutaneous afferents.91 That similar differences are ob-erved in isolated neurons in vitro when exposed to theame inflammatory mediators suggests that these differ-nces do not merely reflect differences in the underlyingissue or the mode of inflammation.89

av1.9 ChannelMany inflammatory mediators act via G-protein–

oupled receptors on voltage-gated sodium channels.ndirect evidence in support of a role for Nav1.9 innflammatory hyperalgesia comes from the observa-ion that guanosine triphosphate (GTP) and its nonhy-rolyzable analog GTP�S results in a significant in-rease in Nav 1.9-mediated current.13 This GTP� S-ediated increase in current is blocked with protein

inase antagonists, suggesting that the increase is theesult of a phosphorylation event. Phosphorylation ofav1.9 is likely to be initiated via inflammatory medi-tors such as PGE2.187 Studies of Nav1.9 currents in
ice lacking the Nav1.8 channel suggest that Nav1.9

conSttGcesdmp

mptidtmsP

AD

pathco

aaceclamidWtm

(ofimctettwp

imfsefcs

MG

indGcmcttttsm

htlnelaT.sfiaat

otb(dtrtdbdrg

edt


hannels may be dynamically modulated by a varietyf factors.152Data on the role of Nav1.9 in the mainte-ance of inflammatory hyperalgesia have been mixed.upport for such a role comes from the observationhat Nav1.9 expression can be modulated by neuro-ropins, in particular, GDNF.78 However, application ofDNF to cultured neurons did not appear to increasehannel expression but rather restored expression lev-ls to normal values. No change in Nav1.9 expression iseen after carrageenan injection,24 but there is evi-ence for a decrease in the amount of Nav1.9-like im-unoreactivity that appears to be transported to theeriphery.48

Mice with nonfunctional Nav1.9 channels have abnor-al responses in inflammatory pain models.174 Theirhase I response in the formalin model is normal, buthey fail to display the phase II response. Following CFAnjection, they develop heat-hyperalgesia to a normalegree for the first few hours, but over the ensuing dayshe hyperalgesia resolves much more quickly than nor-al. Moreover, they do not develop the heat-hyperalge-

ia that is normally seen after subcutaneous injection ofGE2.

ctions of Sodium-Channel Blockers Thato Not Involve Sodium ChannelsAlthough sodium channel actions are undoubtedly therimary site of action for the classical effects of localnesthetics, they are certainly not the sole target ofhese drugs. Interactions with other signaling systemsave been reported for many years, but have not re-eived much attention, because the clinical importancef such effects has never been firmly established.The majority of studies looking at actions not medi-

ted via sodium channels have used lidocaine as the testgent. For example, lidocaine’s interactions with calciumhannels could in theory have an analgesic effect. How-ver, concentrations required for half-maximal block ofalcium channels are between .5 and 2 mmol/L.106 De-ayed rectifier–type potassium channels in Xenopus sci-tic nerve are blocked by lidocaine, but only at approxi-ately 1 mmol/L.29 In anterior pituitary cells, lidocaine

nhibits such potassium currents at 1.9 mmol/L, and so-ium channel inhibition occurs at 170 �mol/L.228

hether disregarding these interactions because ofhese high concentration requirements is appropriate re-ains an open question.Calcium-signaling G protein– coupled receptors

GPCRs), the activation of which results in the releasef calcium from intracellular stores, have been identi-ed as a target for lidocaine.107,113 Several findingsake this discovery of particular interest. First, lido-

aine inhibition of several GPCRs can occur at concen-rations that are clinically relevant, eg, the plasma lev-ls seen after an epidural block. Remarkably, some ofhese receptors are inhibited at nanomolar concentra-ions. Second, it appears that a number of lidocaine’sell described clinical actions that are difficult to ex-
lain by sodium-channel block can be accounted for by t
nhibition of GPCR signaling. In particular, the inflam-atory-modulating actions of lidocaine may result

rom interactions with GPCR-mediated inflammatoryignaling molecules. Importantly, prolonged (hours)xposure of cells to lidocaine greatly enhances its ef-ects on GPCR signaling, whereas the actions on ionhannels take minutes in their onset and to reachteady-state.

olecular Mechanisms of Lidocaine-PCR InteractionsThe mechanisms of action of lidocaine on GPCR signal-

ng have been elucidated largely by studies in recombi-ant systems, usually the Xenopus oocyte. Initial findingsemonstrated a wide range of sensitivities of variousPCR to lidocaine. Thus, when expressed in Xenopus oo-ytes, the M1 muscarinic acetylcholine receptor is re-arkably sensitive to lidocaine, with a 50% inhibition

oncentration (IC50) of 18 nmol/L.108 On the other end ofhe spectrum, the AT1A angiotensin receptor is essen-ially completely insensitive to lidocaine.158 Other recep-or systems, such as those for the inflammatory mediatorhromboxane A2 (TXA) and the putative inflammatoryignaling molecule lysophosphatidate (LPA), show inter-ediate sensitivities to local anesthetics.114,158

Investigations of lidocaine’s effects on GPCR functionave used selective extracellular vs intracellular exposureo QX314 (obtained with bath application and intracel-ular injection, respectively), a nonpermeant perma-ently charged lidocaine analog. Whereas sensitivity toxtracellular QX314 varies greatly (discussed in the fol-owing), the sensitivity to intracellular QX314 is similarmong receptors. For example, IC50 values for QX314 ofXA, LPA, M1, and M3 muscarinic receptors are .53, .72,96, and .45 mmol/L, respectively. Because the moleculartructures of these receptors are quite different, thisnding suggests that intracellular QX314 may be inter-cting instead with similar targets that are downstreamlong the signaling pathways engaged by these recep-ors.The GPCRs activate 1 of several G proteins (Gq, G11, G14,r Go), which in turn activate phospholipase C (PLC). Ac-ivated PLC cleaves membrane phosphatidylinositol-isphosphate into intracellular inositoltrisphosphateIP3) and diacylglycerol (DAG). Inositoltrisphosphate in-uces intracellular calcium release by activating a recep-or channel on intracellular stores. Diacylglycerol is a di-ect PKC activator which works synergistically with Ca�2

o elevate PKC’s activity. In addition to this enzyme-me-iated action, G proteins may directly activate mem-rane channels; an example is the activation of the car-iac inwardly rectifying K� channel by the ��-subuniteleased after muscarinic receptor activation by the va-us nerve.The sensitivity of the various G proteins to local an-

sthetic has been investigated. Calcium release in-uced by direct injection of IP3 into oocytes was foundo be insensitive to local anesthetic,199 indicating that
he IP3 channel and its downstream signaling pathway

aQtofivciLtratswm�

sepcFpdprcbdbt

ackhbsim

mahilegtacbcicctt

ID

lsmflecbitccpcildttesnrei

To

fcawtwieetdmmmtbtd

OR

pfke


re not targets. The similar potencies for intracellularX314 on various receptors, and the lack of effect on

he downstream signaling pathway, suggest an effectn the G proteins that are coupled to receptors. Thesendings have been confirmed by direct knockdown ofarious G protein �-subunits using antisense oligonu-leotides. First, the G protein subtypes coupling to var-ous GPCRs were determined. In Xenopus oocytes, thePA receptor, for example, couples to Gq and Go, therypsin receptor to Gq and G14, and the M1 muscariniceceptor to Gq and G11. Effects of intracellular localnesthetic on receptor function were then studied af-er selective knockdown of each of these G proteinubunits. In each case, the effect of local anestheticas eliminated if, and only if, the Gq protein was re-oved; knockdown of any of the other G protein-subunits was without effect on local anesthetic sen-itivity.113 Interestingly, the lack of local anestheticffect on angiotensin signaling could now be ex-lained also, because this receptor was found not to beoupled to Gq in this model, but instead to Go and G14.urther evidence for an action on Gq signaling wasrovided by studies where mammalian Gq was intro-uced into the cell after knockdown of endogenous Gq

rotein; in this setting, local anesthetic sensitivity wasegained.111 Therefore, the intracellular effects of lo-al anesthetics on GPCR signaling are likely explainedy an inhibitory action on the Gq �-subunit or itsownstream substrate(s). These findings are compati-le with other investigations of local anesthetic ac-ions on GPCR modulation of ion channels.228

It remains possible that the apparent action of localnesthetic on the Gq protein is indirect. For example, it isonceivable that intracellular QX314 activates a proteininase, which in turn phosphorylates Gq and thereby in-ibits its function. Evidence against this hypothesis haseen obtained for PKC by testing for local anestheticensitivity in the presence of a PKC inhibitor,199 but anndirect effect through another kinase cannot be for-

ally excluded.Intracellular inhibition of most GPCRs by QX314,imicking the actions of intracellular lidocaine, occurs

t an IC50 of approximately .5-1 mmol/L, which is muchigher than the nmol/L concentrations required for

nhibition of muscarinic receptor signaling by extracel-ular lidocaine. Which of these sites, intracellular orxtracellular, and which corresponding potency, isermane to the clinical actions of systemic local anes-hetics? Approximately 2-20 �mol/L concentrations arettained systemically in the clinical setting, when lido-aine is given either intentionally intravenously or asyproduct of the epidural administration of high con-entrations and volumes of the local anesthetic. Twomportant observations suggest that these low con-entrations exert effects that might be relevant in thelinical setting: first, synergistic interactions with ex-racellular receptor domains; and second, a profound
ime dependence of the effect. o
nteractions at Extracellular ReceptoromainsIt appears that the high sensitivity of some GPCRs to

ocal anesthetics can be explained by additive or evenuperadditive (synergistic) interactions between 1 (orore) intracellular binding site (of which Gq appears by

ar to be the most prominent) and 1 (or more) extracel-ular site. This has been investigated by comparing theffects of QX314, applied either intracellularly or extra-ellularly, with the effects of the compounds appliedoth intracellularly and extracellularly. When this exper-

ment was performed with the M1 muscarinic receptor,he IC50 values were approximately 50-fold less when theompound was applied to both sides of the membraneompared with selective intracellular or extracellular ap-lication. By isobolographic analysis, there was a signifi-ant superadditivity between the intra- and extracellularnteractions.108 The M3 muscarinic receptor, which isargely similar in structure to the M1 receptor, showsifferent sensitivity: Whereas lidocaine blocks M1 recep-ors with an IC50 of 18 nmol/L, inhibition of the M3 recep-or requires approximately 370 nmol/L.112 This is congru-nt with the presence of a major extracellular bindingite for lidocaine that is present on the M1 receptor butot on the M3 receptor. Studies using chimeric M1/M3

eceptors demonstrated that the N-terminus and thirdxtracellular loop were required to obtain a significantncrease in extracellular inhibition by QX314.

ime Dependence of Lidocaine’s Effectn GPCRThe actions of local anesthetics on GPCR have been

ound to be profoundly time dependent. When oo-ytes expressing TXA receptors were exposed to bupiv-caine for 4 h at 1.2 �mol/L (1/10 of the IC50 obtainedith acute exposure) receptor signaling was reduced

o 25% of control.110 Similar results were obtainedith lidocaine and with other receptors. In general,

nhibitory potency is increased at least 4-fold by 48-hxposure. The site of action of this time-dependentffect was shown to be intracellular and dependent onhe presence of Gq protein: After selective Gq knock-own, time-dependent effects were eliminated. Theolecular mechanism behind this time-dependentodulation of G protein function has not been deter-ined. However, in experiments using GTP�S (a G pro-

ein activator) the inhibition of induced currents byupivacaine is also time-dependent,110 indicating thathe site of action within the G protein pathway isownstream of GDP-GTP exchange.

ther G Proteins May Show Differentesponses to Local AnestheticsWhereas the selective actions on calcium-signaling Groteins have been investigated in some detail, the ef-ects of local anesthetics on other G proteins are not wellnown. Any such effect is likely to be different from theffect shown on the Gq pathway. Local anesthetic effects
n Gi have been examined in a reconstituted adenosine

sGispa

LS

nmssctTrnlt

E

rnadkGir

itt

LP

navnshhi

thvwiIp�ttfimpwii

C

opltmdClmCalpcncnauata

Fpniec.toam


ignaling system.15 Lidocaine was found to potentiate

i-coupled signaling through the A1 receptor by reduc-ng cyclic adenosine monophosphate production. Gi washown to be the target, indicating that in this class of Groteins lidocaine may have an enhancing, rather thann inhibiting, receptor-signaling action.

ocal Anesthetic Effects on the Immuneystem Response to Tissue DamageThe pain associated with tissue inflammation anderve injury occurs in the context of a complex im-une response. The immune response is likely to make

ignificant contributions to nociceptor activation andensitization, and to the generation of ectopic dis-harge, for example, via the release of proinflamma-ory cytokines (for review, see Watkins and Maier216).herefore, a local anesthetic effect on the immuneesponse is likely to modify both inflammatory andeuropathic pain states. There is now evidence that

ocal anesthetics do affect the immune response andhat this occurs at low plasma levels.

ffects on Human NeutrophilsThe immune response to tissue injury is initiated by the

apid infiltration of neutrophils. Human neutrophils doot express sodium channels, and any actions of localnesthetic on these cells are therefore by definition so-ium-channel independent. Neutrophil functioning isnown to be regulated by GPCR signaling; therefore, thePCR actions discussed above may also be relevant to the

mmune system. Neutrophils have a complex array of

igure 10. Effects of lidocaine (10�7 mol/L) on priming of su-eroxide release from human neutrophils in vitro. (1) Primingeutrophils with an exposure to PAF and subsequently activat-

ng them with formyl-methionine-leucine-phenylalanine (fMLP)vokes a large release of superoxide anion. (2) Including lido-aine with PAF in the priming step significantly decreases (*P �01) the neutrophil response to subsequent activation (3 and 4;he two curves superimpose) Activating neutrophils with fMLPr fMLP � lidocaine evokes a much smaller response in thebsence of priming (ie, prior exposure to PAF). Data from Holl-ann et al.109

esponses to activation by mediators, including superox- o

de release, chemotaxis, and release of further media-ors; local anesthetics have been shown to inhibit all ofhese.76,201

ocal Anesthetics Inhibit Neutrophilriming but Not ActivationAn area that has been investigated in some detail iseutrophil priming. Priming refers to a process wheren initial exposure to an agonist, such as platelet-acti-ating factor (PAF), induces little or no response fromeutrophils but results in a major increase in their re-ponse to a subsequent activating agonist.109 The en-anced release of superoxide from primed neutrophilsas been shown to be an important pathogenic factor

n a number of inflammatory diseases.49

The effects of various local anesthetics on the activa-ion and priming pathways have been investigated inuman neutrophils in vitro (Fig 10). At clinically rele-ant concentrations, local anesthetics are essentiallyithout effect on the activating pathway; but priming

s inhibited substantially by even very low levels.76,109

mportantly, the effect on priming in human neutro-hils is strongly time dependent: Lidocaine at .1mol/L inhibited PAF-induced priming to 39% of con-rol after a 5-hour exposure.110 The mechanism behindhe selective action on priming can be explained by thendings already discussed. Neutrophil activation isediated primarily by a Gi-coupled pathway, whereas

riming is mediated primarily by a Gq-coupled path-ay. A selective effect of local anesthetic on Gq signal-

ng explains why activation is unaffected while prim-ng is inhibited.

onclusionsIt is clear that changes in the expression and functionf voltage-gated sodium channel isoforms play an im-ortant role in neuropathic pain. These changes are

ikely to be key factors in the pathogenesis of both spon-aneous and evoked ectopic discharge in damaged pri-ary afferent neurons and probably also in the ectopic

ischarge that develops in their undamaged neighbors.hanges in channel isoforms may also produce patho-

ogic changes in the process that regulates neurotrans-itter release. Changes in sodium channel expression in

NS neurons, leading to altered states of neuronal excit-bility, are also likely to be involved. Moreover, accumu-ating evidence points to changes in sodium channel ex-ression as key factors in inflammatory pain states. Suchhanges include many of the phenomena implicated ineuropathic pain, with the additional factor of a role forhanges in sodium channel expression in the process ofociceptor sensitization. There are data to suggest thatlterations in sodium channel expression may be partic-larly important in chronic inflammatory pain, and this isn area that deserves far more investigation. Whetherhe key change is in a single channel isoform or whetherparticular mix of channels is the change in neuropathic
r inflammatory pain states remains to be determined,

bs

pwsdwsc

cno

Hdtctt

acflmo

R

1is

2ULdw

3p7

4r

5p5

6i

7lt1

8sm1

9tt

1rl2

1Aip

1m2

1


ut a change in the “cocktail” of channels is likely to beignificant.At least some neuropathic pain phenomena, anderhaps chronic inflammatory pain phenomena asell, are unusually sensitive to local anesthetics. This

ensitivity is likely to involve the drugs’ action on so-ium channels in some cases (eg, ectopic discharge),hereas other cases (eg, the long-duration pain relief

een after a brief infusion) may involve other pharma-ologic pathways.The clinical efficacy of lidocaine and other sodium-

hannel blockers for the control of inflammatory andeuropathic pain is very likely to include mechanisms
f action that involve binding to the sodium channel. d
2. Arner S, Lindblom U, Meyerson BA, Molander C: Pro-

lcP

1Ger

1Hhch2

1mA

1(s

1tm

1TgaG

1McjB

2an

2svr

2ddi

2

owever, there is accumulating evidence that theserugs may have a clinically significant pharmacologyhat does not involve their binding to the sodiumhannel. Potent effects on G protein– coupled recep-ors and on the neuroimmune interactions that con-ribute to pain hypersensitivity states may be involved.Local anesthetics such as lidocaine and bupivacaine (and

lmost certainly others) have a remarkably broad pharma-ology. Potentially clinically relevant actions have beenound with nmol/L to mmol/L concentrations. It is particu-arly noteworthy that some of these drug effects are clearly

anifest only with long-term exposure, a factor that is notften investigated experimentally. Clearly there is a great
eal that is new about these old drugs.194,235
eferences

. Abram SE, Yaksh TL: Systemic lidocaine blocks nervenjury–induced hyperalgesia and nociceptor-driven spinalensitization in the rat. Anesthesiology 80:383-391, 1994

. Akopian AN, Souslova V, England S, Okuse K, Ogata N,re J, Smith A, Kerr BJ, McMahon SB, Boyce S, Hill R, StanfaC, Dickenson AH, Wood JH: The tetrodotoxin-resistant so-ium channel SNS has a specialized function in pain path-ays. Nat Neurosci 2:541-548, 1999

. Aley KO, McCarter G, Levine JD: Nitric oxide signaling inain and nociceptor sensitization in the rat. J Neurosci 18:008-7014, 1998

. Amir R, Devor M: Chemically mediated cross-excitation inat dorsal root ganglia. J Neurosci 16:4733-4741, 1996

. Amir R, Devor M: Spike-evoked suppression and burstatterning in dorsal root ganglion neurons. J Physiol (Lond)01:183-196, 1997

. Amir R, Liu CN, Kocsis JD, Devor M: Oscillatory mechanismn primary sensory neurons. Brain 125:421-435, 2002

. Amir R, Michaelis M, Devor M: Membrane potential oscil-ations in dorsal root ganglion neurons: Role in normal elec-rogenesis and neuropathic pain. J Neurosci 19:8589-8596,999

. Amir R, Michaelis M, Devor M: Burst discharge in primaryensory neurons: triggered by subthreshold oscillations,aintained by depolarizing afterpotentials. J Neurosci 22:

187-1198, 2002

. Andrew D, Greenspan JD: Mechanical and heat sensitiza-ion of cutaneous nociceptors after peripheral inflamma-ion in the rat. J Neurophysiol 82:2649-2656, 1999

0. Araujo MC, Sinnott CJ, Strichartz GR: Multiple phases ofelief from experimental mechanical allodynia by systemicidocaine: Responses to early and late infusions. Pain 103:21-9, 2003

1. Argoff CE, Galer BS, Jensen MP, Oleka N, GammaitoniR: Effectiveness of the lidocaine patch 5% on pain qualities

n three chronic pain states: Assessment with the Neuro-athic Pain Scale. Curr Med Res Opin 20Suppl 2S21-S28, 2004

1b. Armstrong CM, Bezanilla F: Currents related to move-ents of the gating particles of the sodium channels. Nature

42:459-461, 1973

onged relief of neuralgia after regional anesthetic blocks. Aall for further experimental and systematic clinical studies.ain 43:287-297, 1990

3. Baker MD, Chandra SY, Ding Y, Waxman SG, Wood JN:TP-induced tetrodotoxin-resistant Na� current regulatesxcitability in mouse and rat small diameter sensory neu-ones. J Physiol 548:373-382, 2003

4. Bender RA, Soleymani SV, Brewster AL, Nguyen ST, Beck, Mathern GW, Baram TZ: Enhanced expression of a specificyperpolarization-activated cyclic nucleotide–gated cationhannel (HCN) in surviving dentate gyrus granule cells ofuman and experimental epileptic hippocampus. J Neurosci3:6826-6836, 2003

5. Benkwitz C, Garrison JC, Linden J, Durieux ME, Holl-ann MW: Lidocaine enhances Galphai protein function.nesthesiology 99:1093-1101, 2003

6. Bennett GJ: Neuropathic pain, in Wall PD, Melzack Reds): Textbook of Pain. Edinburgh, UK, Churchill Living-tone, 1994, pp 201-224

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

8. Beyak MJ, Ramji N, Krol KM, Kawaja MD, Vanner SJ: TwoTX-resistant Na� currents in mouse colonic dorsal rootanglia neurons and their role in colitis-induced hyperexcit-bility. Am J Physiol Gastrointest Liver Physiol 287:G845-855, 2004

9. Bian D, Ossipov MH, Ibrahim M, Raffa RB, Tallarida RJ,alan TP, Lai J, Porreca, F: Loss of antiallodynic and antino-

iceptive spinal/supraspinal morphine synergy in nerve-in-ured rats: Restoration by MK-801 or dynorphin antiserum.rain Res 831:55-63, 1999

0. Bielefeldt K, Ozaki N, Gebhart GF: Experimental ulcerslter voltage-sensitive sodium currents in rat gastric sensoryeurons. Gastroenterology 122:394-405, 2002

1. Bischoff U, Brau ME, Vogel W, Hempelmann G, Ol-chewski A. Local anesthetics block hyperpolarization-acti-ated inward current in rat small dorsal root ganglion neu-ones. Br J Pharmacol 139:1273-80, 2003

2. Black JA, Cummins TR, Plumpton C, Chen YH, Hormuz-iar W, Clare JJ, Waxman SG: Upregulation of a silent so-ium channel after peripheral, but not central, nerve injury

n DRG neurons. J Neurophysiol 82:2776-2785, 1999

3. Black JAS, Dib-Hajj S, McNabola K, Jeste S, Rizzo MA,

Km4

2Cct

2cB

2t

2Mp

2eas

2etn

3se

3ns4

3pp

3Tfn2

3l

3nB

3cN

3rds

3cmP

3nc

3Tsr

4na3

4an

4mpp

4mi

4Ldal

4np

4mc1

4p

4sN2

4ia

5Top

5mm

5Jss

5lp

5pm

5goA

51

5o


ocsis JD, Waxman SG: Spinal sensory neurons expressultiple sodium channel �-subunit mRNAs. Mol Brain Res

3:117-131, 1996

4. Black JA, Liu S, Tanaka M, Cummins TR, Waxman SG:hanges in the expression of tetrodotoxin-sensitive sodiumhannels within dorsal root ganglia neurons in inflamma-ory pain. Pain 108:237-247, 2004

5. Blumenthal KM, Seibert AL: Voltage-gated sodiumhannel toxins: Poisons, probes, and future promise. Celliochem Biophys 38:215-238, 2003

6. Boas RA, Covino BG, Shahnarian A: Analgesic responseso i.v. lignocaine. Br J Anaesth 54:501-504, 1982

7. Boucher TJ, Okuse K, Bennett DL, Munson JB, Wood JN,cMahon SB: Potent analgesic effects of GDNF in neuro-

athic pain states. Science 290:124-127, 2000

8. Brau ME, Dreimann M, Olschewski A, Vogel W, Hemp-lmann G: Effect of drugs used for neuropathic pain man-gement on tetrodotoxin-resistant Na(�) currents in ratensory neurons. Anesthesiology 94:137-144, 2001

9. Brau ME, Vogel W, Hempelmann G: Fundamental prop-rties of local anesthetics: Half-maximal blocking concentra-ions for tonic block of Na� and K� channels in peripheralerve. Anesth Analg 87:885-889, 1998

0. Brock JA, McLachlan EM, Belmonte C: Tetrodotoxin-re-istant impulses in single nociceptor nerve terminals in guin-a-pig cornea. J Physiol (Lond) 512:211-217, 1998

1. Brock JA, Pianova S, Belmonte C: Differences betweenerve terminal impulses of polymodal nociceptors and coldensory receptors of the guinea-pig cornea. J Physiol 533:93-501, 2001

2. Bryson HM, Wilde MI: Amitriptyline. A review of itsharmacological properties and therapeutic use in chronicain states. Drugs Aging 8:459-476, 1996

3. Burgess SE, Gardell LR, Ossipov MH, Malan TP, VanderahW, Lai J, Porreca F: Time-dependent descending facilitationrom the rostral ventromedial medulla maintains, but doesot initiate, neuropathic pain. J Neurosci 22:5129-5136,002

4. Butterworth JF, Strichartz GR: Molecular mechanisms ofocal anesthesia: A review. Anesthesiology 72:711-734, 1990

4b. Cahalan MD, Almers W: Interactions between quater-ary lidocaine, the sodium channel gates, and tetrodotoxin.iophys J 27:39-55, 1979

5. Cantrell AR, Catterall WA: Neuromodulation of Na�hannels: An unexpected form of cellular plasticity. Nat Reveurosci 2:397-407, 2001

6. Cardenas CG, Del Mar LP, Cooper BY, Scroggs RS: 5HT4eceptors couple positively to tetrodotoxin-insensitive so-ium channels in a subpopulation of capsaicin-sensitive ratensory neurons. J Neurosci 17:7181-7189, 1997

7. Carr RW, Pianova S, Brock JA: The effects of polarizingurrent on nerve terminal impulses recorded from poly-odal and cold receptors in the guinea-pig cornea. J Gen

hysiol 120:395-405, 2002

8. Catterall WA: From ionic currents to molecular mecha-isms: The structure and function of voltage-gated sodiumhannels. Neuron 26:13-25, 2000

9. Casula MA, Facer P. Powell AJ, Kinghorn IJ, Plumpton C,ate SN, Bountra C, Birch R, Anand P: Expression of theodium channel �3 subunit in injured human sensory neu-
ons. Neuroreport 15:1629-1632, 2004 N
0. Chabal C, Russell LC, Burchiel KJ: The effect of intrave-ous lidocaine, tocainide, and mexiletine on spontaneouslyctive fibers originating in rat sciatic neuromas. Pain 38:333-38, 1989

1. Chaplan SR, Bach FW, Shafer SL, Yaksh TL: Prolongedlleviation of tactile allodynia by intravenous lidocaine ineuropathic rats. Anesthesiology 83:775-785, 1995

2. Chaplan SR, Guo HQ, Lee DH, Luo L, Liu C, Kuei C, Velu-ain AA, Butler MP, Brown SM, Dubin AE. Neuronal hyper-olarization-activated pacemaker channels drive neuro-athic pain. J Neurosci 23:1169-1178, 2003

3. Chapman V, Ng J, Dickenson AH: A novel spinal action ofexiletine in spinal somatosensory transmission of nerve

njured rats. Pain 77:289-296, 1998

4. Chen Q, King T, Vanderah TW, Ossipov MH, Malan TP,ai J, Porreca F: Differential blockade of nerve injury–in-uced thermal and tactile hypersensitivity by systemicallydministered brain-penetrating and peripherally restricted

ocal anesthetics. J Pain 5:281-289, 2004

5. Chernoff DM: Kinetic analysis of phasic inhibition ofeuronal sodium currents by lidocaine and bupivacaine. Bio-hys J 58:53-68, 1990

6. Chevrier P, Vijayaragavan K, Chahine M: Differentialodulation of Nav1.7 and Nav1.8 peripheral nerve sodium

hannels by the local anesthetic lidocaine. Br J Pharmacol42:576-584, 2004

7. Chung JM, Chung K: Sodium channels and neuropathicain. Novartis Found Symp 261:27-31, 2004

8. Coggeshall RE, Tate S, Carlton SM: Differential expres-ion of tetrodotoxin-resistant sodium channels Nav1.8 andav1.9 in normal and inflamed rats. Neurosci Lett 355:45-48,004

9. Condliffe AM, Kitchen E, Chilvers ER: Neutrophil prim-ng: Pathophysiological consequences and underlying mech-nisms. Clin Sci (Lond) 94:461-471, 1998

0. Coward K, Mosahebi A, Plumpton C, Facer P, Birch R,ate S, Bountra C, Terenghi G, Anand P: Immunolocalisationf sodium channel NaG in the intact and injured humaneripheral nervous system. J Anat 198:175-80, 2001

1. Cranner MJ, Klein JP, Renganathan M, Black JA, Wax-an SG: Changes of sodium channel expression in experi-ental diabetic neuropathy. Ann Neurol 52:786-792, 2002

2. Cummins TR, Dib-Hajj SD, Black JA, Akopian AN, WoodN, Waxman SG: A novel persistent tetrodotoxin-resistantodium current in SNS-null and wild-type small primary sen-ory neurons. J Neurosci 19:RC43, 1999

3. Cummins TR, Dib-Hajj SD, Waxman SG: Electrophysio-ogical properties of mutant Nav1.7 sodium channels in aainful inherited neuropathy. J Neurosci 24:8232-8236, 2004

4. Dalpiaz AS, Lordon SP, Lipman AG: Topical lidocaineatch therapy for myofascial pain. J Pain Palliat Care Phar-acother 18:15-34, 2004

5. Davis MD, Sandroni P, Rooke TW, Low PA. Erythromelal-ia: Vasculopathy, neuropathy, or both? A prospective studyf vascular and neurophysiologic studies in erythromelalgia.rch Dermatol 139:1337-1343, 2003

6. de Jong RH: Local Anesthetics. St Louis, MO, Mosby,994, pp 250-258

7. Decosterd I, Ji RR, Abdi S, Tate S, Woolf CJ: The patternf expression of the voltage-gated sodium channels
a(v)1.8 and Na(v)1.9 does not change in uninjured primary

sP

5e

5icJ

6ib1

6ec

6NN

6Lni

6Sto

6nia

6tn

6Spn2

6doa

6Jan

7Yst

7trc

7ei

7mrm

7ao

7kJp1

7Gth

7cs5

7mnn

7Ain

7aq1

8Alc

8muO

8mBstt

8Aei1

8c2

8lt

8is

8Iz1

8g


ensory neurons in experimental neuropathic pain models.ain 96:269-277, 2002

8. Devor M: Neuropathic pain and injured nerve: Periph-ral mechanisms. Br Med Bull 47:619-630, 1991

9. Devor M, Govrin-Lippmann R, Angelides K: Na� channelmmunolocalization in peripheral mammalian axons andhanges following nerve injury and neuroma formation.Neurosci 13:1976-1992, 1993

0. Devor M, Seltzer Z: Pathophysiology of damaged nervesn relation to chronic pain, in Wall PD, Melzack R (eds): Text-ook of Pain. London, UK, Churchill-Livingstone, 1999, pp29-164

1. Devor M, Wall PD, Catalan N: Systemic lidocaine silencesctopic neuroma and DRG discharge without blocking nerveonduction. Pain 48:261-268, 1992

2. Dib-Hajj S, Black JA, Cummins TR, Waxman SG: NaN/av1.9: A sodium channel with unique properties. Trendseurosci 25: 253-259, 2002

3. Dib-Hajj S, Fjell J, Cummins TR, Zheng Z, Fried K,aMotte R, Black JA, Waxman SG: Plasticity of sodium chan-el expression in DRG neurons in the chronic constriction

njury model of neuropathic pain. Pain 83;591-600, 1999

4. Dib-Hajj SD, Rush AM, Cummins TR, Hisama FM, Novella, Tyrrell L, Marshall L, Waxman SG: Gain-of-function muta-ion in Nav1.7 in familial erythromelalgia induces burstingf sensory neurons. Brain 128:1847-1854, 2005

5. Dib-Hajj SD, Tyrrell L, Black JA, Waxman SG: NaN, aovel voltage-gated Na channel, is expressed preferentially

n peripheral sensory neurons and down-regulated afterxotomy. Proc Natl Acad Sci U S A 95:8963-8968, 1998

6. Dina, OA, McCarter GC, de Coupade C, Levine JD: Role ofhe sensory neuron cytoskeleton in second messenger sig-aling for inflammatory pain. Neuron 39:613-624, 2003

7. Djouhri L, Fang X, Okuse K, Wood JN, Berry CM, LawsonN: The TTX-resistant sodium channel Nav1.8 (SNS/PN3): ex-ression and correlation with membrane properties in ratociceptive primary afferent neurons. J Physiol 550:739-752,003

8. Doan TN, Stephans K, Ramirez AN, Glazebrook PA, An-resen MC, Kunze DL: Differential distribution and functionf hyperpolarization-activated channels in sensory neuronsnd mechanosensitive fibers. J Neurosci 24:3335-3343, 2004

9. Drenth JP, te Morsche RH, Guillet G, Taieb A, Kirby RL,ansen JB: SCN9A mutations define primary erythermalgias a neuropathic disorder of voltage-gated sodium chan-els. J Invest Dermatol 124:1333-1338, 2005

0. Duarte AM, Pospisilova E, Reilly E, Mujenda F, Hamaya, Strichartz GR: Reduction of post-incisional allodynia byubcutaneous bupivacaine: Findings with a new model inhe hairy skin of the rat. Anesthesiol 103:113-125, 2005

1. England S, Bevan S, Docherty RJ: PGE2 modulates theetrodotoxin-resistant sodium current in neonatal rat dorsaloot ganglion neurons via the cyclic AMP–protein kinase Aascade. J Physiol 495:429-440, 1996

2. Eschenfelder S, Habler HJ, Janig W: Dorsal root sectionlicits signs of neuropathic pain rather than reversing themn rats with L5 spinal nerve injury. Pain 87:213-219, 2000

3. Fang X, Djouhri L, Black JA, Dib-Hajj S, Lawson SN, Wax-an SG: The presence and the role of the tetrodotoxin-

esistant sodium channel Na 1.9 (NaN) in nociceptive pri-
vary afferent neurons. J Neurosci 22:7425-7433, 2002 i
4. Ferrante FM, Paggiol J, Cherukuri S, Arthur GR: The an-lgesic response to intravenous lidocaine in the treatmentf neuropathic pain. Anesth Analg 82:91-97, 1996

5. Finnerup NB, Biering-Sorensen F, Johannesen H, Ter-elsen AJ, Juhl GI, Kristensen AD, Sindrup SH, Bach FW,ensen TS. Intravenous lidocaine relieves spinal cord injuryain: A randomzed controlled trial. Anesthesiol 102:1023-030, 2005

6. Fischer LG, Bremer M, Coleman EJ, Conrad B, Krumm B,ross A, Hollmann MW, Mandell G, Durieux ME: Local anes-

hetics attenuate lysophosphatidic acid–induced priming inuman neutrophils. Anesth Analg 92:1041-1047, 2001

7. Fitzgerald EM, Okuse K, Wood JN, Dolphin AC, Moss SJ:AMP-dependent phosphorylation of the tetrodotoxin-re-istant voltage-dependent sodium channel SNS. J Physiol16:433-446, 1999

8. Fjell J, Cummins TR, Dib-Hajj SD, Fried K, Black JA, Wax-an SG: Differential role of GDNF and NGF in the mainte-ance of two TTX-resistant sodium channels in adult DRGeurons. Brain Res Mol Brain Res 67:267-282, 1999

9. Fjell J, Hjelmstrom P, Hormuzdiar W, Milenkovic M,glieco F, Tyrrell L, Dib-Hajj S, Waxman SG, Black JA: Local-

zation of the tetrodotoxin-resistant sodium channel NaN inociceptors. Neuroreport 11:199-202, 2000

9b. Frazier DT, Narahashi T, Yamada M: The site of actionnd active form of local anesthetics. II. Experiments withuaternary compounds. J Pharmacol Exp Ther 171:45-51,970

0. Fridrich P, Eappen S, Jaeger W, Schernhammer E, ZizzaM, Wang GK, Gerner P: Phase Ia and Ib study of amitripty-

ine for ulnar nerve block in humans: side effects and effi-acy. Anesthesiology 100:1511-1518, 2004

1. Galer BS, Sheldon E, Patel N, Codding C, Burch F, Gam-aitoni AR: Topical lidocaine patch 5% may target a novelnderlying pain mechanism in osteoarthritis. Curr Med Respin 20:1455-1458, 2004

2. Gardell LR, Wang R, Ehrenfels C, Ossipov MH, Rosso-ando AJ, Miller S, Buckley C, Cai AK, Tse A, Foley SF, Gong, Walus L, Carmillo P, Worley D, Huang C, Engber T, Pepin-ky B, Cate RL, Vanderah TW, Lai J, Sah DW, Porreca F: Mul-iple actions of systemic artemin in experimental neuropa-hy. Nat Med 9:1383-1389, 2003

3. Gerner P, Haderer AE, Mujtaba M, Sudoh Y, Narang S,bdi S, Srinivasa V, Pertl C, Wang GK: Assessment of differ-ntial blockade by amitriptyline and its N-methyl derivative

n different species by different routes. Anesthesiology 98:484-1490, 2003

4. Gerner P, Kao G, Srinivasa V, Narang S, Wang GK: Topi-al amitriptyline in healthy volunteers. Reg Anesth Pain Med8:289-293, 2003

5. Gerner P, Mujtaba M, Sinnott CJ, Wang GK: Amitripty-ine versus bupivacaine in rat sciatic nerve blockade. Anes-hesiology 94:661-667, 2001

6. Gold MS, Levine JD: DAMGO inhibits prostaglandin E2-nduced potentiation of a TTX-resistant Na� current in ratensory neurons in vitro. Neurosci Lett 212:83-86, 1996

7. Gold MS, Levine JD, Correa AM: Modulation of TTX-RNa by PKC and PKA and their role in PGE2-induced sensiti-ation of rat sensory neurons in vitro. J Neurosci 18:10345-0355, 1998

8. Gold MS, Reichling DB, Shuster MJ, Levine JD: Hyperal-esic agents increase a tetrodotoxin-resistant Na� current

n nociceptors. Proc Natl Acad Sci U S A 93:1108-1112, 1996

8rct

9re

9EJ

9A

9HNN2

9si7

9LNF5

9Lic

9Str

9Ar

9et

1mtwr

1Asi

1ps2

1d2

1rtJ

1

m1

1an1

1Ga

1Ls

1at

1Sia

1It

1cl2

1LsM

1rbo

1Eert

1c1

1ov

1ta2

1naa

1ca1

1lr


9. Gold MS, Traub RJ: Cutaneous and colonic rat DRG neu-ons differ with respect to both baseline and PGE2-inducedhanges in passive and active electrophysiological proper-ies. J Neurophysiol 91:2524-2531, 2004

0. Gold MS, Weinreich D, Kim CS, Wang R, Treanor J, Por-eca F, Lai J: Redistribution of Na(V)1.8 in uninjured axonsnables neuropathic pain. J Neurosci 23:158-166, 2003

1. Gold MS, Zhang L, Wrigley DL, Traub RJ: Prostaglandin(2) modulates TTX-R I(Na) in rat colonic sensory neurons.Neurophysiol 88:1512-1522, 2002

2. Goldin AL: Resurgence of sodium channel research.nnu Rev Physiol 63:871-894, 2001

3. Goldin AL, Barchi RL, Caldwell JH, Hofmann F, Howe JR,unter JC, Kallen RG, Mandel G, Meisler MH, Netter YB,oda M, Tamkun NM, Waxman SG, Wood JN, Catterall WA:omenclature of voltage-gated sodium channels. Neuron8:365-368, 2000

4. Gould HJ 3rd, England JD, Liu ZP, Levinson SR: Rapidodium channel augmentation in response to inflammationnduced by complete Freund’s adjuvant. Brain Res 802:69-4, 1998

5. Gould HJ 3rd, England JD, Soignier RD, Nolan P, MinorD, Liu ZP, Levinson SR, Paul D: Ibuprofen blocks changes ina v 1.7 and 1.8 sodium channels associated with completereund’s adjuvant–induced inflammation in rat. J Pain:270-280, 2004

6. Gould HJ 3rd, Gould TN, England JD, Paul D, Liu ZP,evinson SR, Reeb SC: A possible role for nerve growth factorn the augmentation of sodium channels in models ofhronic pain. Brain Res 854:19-29, 2000

7. Gould HJ 3rd, Gould TN, Paul D, England JD, Liu ZP, ReebC, Levinson SR: Development of inflammatory hypersensi-ivity and augmentation of sodium channels in rat dorsaloot ganglia. Brain Res 824:296-299, 1999

8. Gracely RH, Lynch S, Bennett GJ. Painful neuropathy:ltered central processing, maintained dynamically by pe-

ipheral input. Pain 51:175-194, 1992

9. Gu JG, MacDermott AB: Activation of ATP P2X receptorslicits glutamate release from sensory neuron synapses. Na-ure 389:749-753, 1997

00. Hains BC, Klein JP, Saab CY, Craner MJ, Black JA, Wax-an SG: Upregulation of sodium channel Nav1.3 and func-

ional involvement in neuronal hyperexcitability associatedith central neuropathic pain after spinal cord injury. J Neu-

osci 23:8881-8892, 2003

01. Hains BC, Saab CY, Klein JP, Craner MJ, Waxman SG:ltered sodium channel expression in second-order spinal

ensory neurons contributes to pain after peripheral nervenjury. J Neurosci 24:4832-4839, 2004

02. Hains BC, Saab CY, Waxman SG: Changes in electro-hysiological properties and sodium channel Nav1.3 expres-ion in thalamic neurons after spinal cord injury. Brain 128:359-2371, 2005

03. Han HC, Lee DH, Chung JM: Characteristics of ectopicischarges in a rat neuropathic pain model. Pain 84:253-261,000

04. Herzog RI, Cummins TR, Waxman SG: Persistent TTX-esistant Na� current affects resting potential and responseo depolarization in simulated spinal sensory neurons.Neurophysiol 86:1351-1364, 2001

05. Hille B. The permeability of the sodium channel to 1

etal cations in myelinated nerve. J Gen Physiol 59: 637-658,972

06. Hirota K, Browne T, Appadu BL, Lambert DG: Do localnaesthetics interact with dihydropyridine binding sites oneuronal L-type Ca2� channels? Br J Anaesth 78:185-188,997

07. Hollmann MW, Difazio CA, Durieux ME: Ca-signaling-protein–coupled receptors: A new site of local anestheticction? Reg Anesth Pain Med 26:565-571, 2001

08. Hollmann MW, Fischer LG, Byford AM, Durieux ME:ocal anesthetic inhibition of m1 muscarinic acetylcholineignaling. Anesthesiology 93:497-509, 2000

09. Hollmann MW, Gross A, Jelacin N, Durieux ME: Localnesthetic effects on priming and activation of human neu-rophils. Anesthesiology 95:113-122, 2001b

10. Hollmann MW, Herroeder S, Kurz KS, Hoenmann CW,truemper D, Hahnenkamp K, Durieux ME: Time-dependentnhibition of G protein–coupled receptor signaling by localnesthetics. Anesthesiology 100:852-860, 2004

11. Hollmann MW, McIntire WE, Garrison JC, Durieux ME:nhibition of mammalian Gq protein function by local anes-hetics. Anesthesiology 97:1451-1457, 2002

12. Hollmann MW, Ritter CH, Henle P, de Klaver M, Kamat-hi GL, Durieux ME: Inhibition of m3 muscarinic acetylcho-ine receptors by local anaesthetics. Br J Pharmacol 133:207-16, 2001c

13. Hollmann MW, Wieczorek KS, Berger A, Durieux ME:ocal anesthetic inhibition of G protein–coupled receptorignaling by interference with Galpha(q) protein function.ol Pharmacol 59:294-301, 2001d

14. Honemann CW, Arledge JA, Podranski T, Aken HV, Du-ieux M: Volatile and local anesthetics interfere with throm-oxane A2 receptors recombinantly expressed in Xenopusocytes. Adv Exp Med Biol 469:277-283, 1999

15. Hong S, Morrow TJ, Paulson PE, Isom LL, Wiley JW:arly painful diabetic neuropathy is associated with differ-ntial changes in tetrodotoxin-sensitive and tetrodotoxin-esistant sodium channels in dorsal root ganglion neurons inhe rat. J Biol Chem 279:29341-29350, 2004

16. Jeftinija S: The role of tetrodotoxin-resistant sodiumhannels of small primary afferent fibers. Brain Res 639:125-34, 1994

17. Jett MF, McGuirk J, Waligora D, Hunter JC: The effectsf mexiletine, desipramine and fluoxetine in rat models in-olving central sensitization. Pain 69:161-169, 1997

18. Ji RR, Woolf CJ: Neuronal plasticity and signal transduc-ion in nociceptive neurons: Implications for the initiationnd maintenance of pathological pain. Neurobiol Dis 8:1-10,001

19. Kajander KC, Bennett GJ: Onset of a painful peripheraleuropathy in rat: A partial and differential deafferentationnd spontaneous discharge in A beta and A delta primaryfferent neurons. J Neurophysiol 68:734-744, 1992

20. Kajander KC, Wakisaka S, Bennett GJ: Spontaneous dis-harge originates in the dorsal root ganglion at the onset ofpainful peripheral neuropathy in the rat. Neurosci Lett

38:225-228, 1992

21. Kalso E, Tramer MR, McQuay HJ, Moore RA: Systemicocal-anaesthetic-type drugs in chronic pain: A systematiceview. Eur J Pain 2:3-14, 1998

22. Kapoor R, Li Y-G, Smith KJ: Slow sodium-dependent

pm

1Ip

1ola1

1th3

1ti

1l1

1tN

1dpJ

1dn2

1enM

1et

1mgB

1rn1

1Psp

1in

1PpN

1ac1

1vn

1Bro6

1r1

1sp

1omp

1itr

1MAn

1SL

1tm

1S

1Lwi

1r

1Mcm

1ip

1NtNA

1Rp1

1sP

1


otential oscillations contribute to ectopic firing inammalian demyelinated axons. Brain 120:647-652, 1997

23. Kastrup J, Petersen P, Dejgard A, Angelo HR, Hilsted J:ntravenous lidocaine infusion—a new treatment of chronicainful diabetic neuropathy? Pain 28:69-75, 1987

24. Kawai R, Fujita S, Suzuki T. Simultaneous quantitationf lidocaine and its four metabolites by high-performance

iquid chromatography: Application to studies on in vitrond in vivo metabolism of lidocaine in rats. J Pharm Sci 74:219-1224, 1985

25. Kerr BJ, Souslova V, McMahon SB, Wood JN: A role forhe TTX-resistant sodium channel Nav 1.8 in NGF-inducedyperalgesia, but not neuropathic pain Neuroreport 12:077-3080, 2001

25b. Keynes RD, Rojas E: Kinetics and steady-state proper-ies of the charged system controlling sodium conductancen the squid giant axon. J Physiol 239:393-434, 1974

26. Khan MA, Gerner P, Wang GK: Amitriptyline for pro-onged cutaneous analgesia in the rat. Anesthesiology 96:09-116, 2002

27. Khasar SG, Gold MS, Levine JD: A tetrodotoxin-resis-ant sodium current mediates inflammatory pain in the rat.eurosci Lett 256:17-20, 1998

28. Khasar SG, McCarter G, Levine JD: Epinephrine pro-uces a beta-adrenergic receptor-mediated mechanical hy-eralgesia and in vitro sensitization of rat nociceptors.Neurophysiol 81:1104-1112, 1999

29. Khodorova A, Meissner K, LeeSon S, Strichartz GR: Li-ocaine selectively blocks abnormal impulses arising fromoninactivating Na channels. Muscle Nerve 24:634-647,001

30. Kim CH, Oh Y, Chung JM, Chung K: The changes inxpression of three subtypes of TTX-sensitive sodium chan-els in sensory neurons after spinal nerve ligation. Brain Resol Brain Res 95:153-161, 2001

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

32. Koltzenburg M., Torebjork HE, Wahren LK: Nociceptorodulated central sensitization causes mechanical hyperal-

esia in acute chemogenic and chronic neuropathic pain.rain 117Pt 3579-591, 1994

33. Kostyuk PG, Veselovsky NS, Tsyndrenko AY: Ionic cur-ents in the somatic membrane of rat dorsal root ganglioneurons—I. Sodium currents. Neuroscience 6:2423-2430,981

34. Kovelowski CJ, Ossipov MH, Sun H, Lai J, Malan TP,orreca F: Supraspinal cholecystokinin may drive tonic de-cending facilitation mechanisms to maintain neuropathicain in the rat. Pain 87:265-273, 2000.

35. Kral MG, Xiong Z, Study RE: Alteration of Na� currentsn dorsal root ganglion neurons from rats with a painfuleuropathy. Pain 81:15-24, 1999

36. Lai J, Gold MS, Kim CS, Bian D, Ossipov MH, Hunter JC,orreca F: Inhibitions of neuropathic pain by decreased ex-ression of the tetrodotoxin-resistant sodium channel,aV1.8. Pain 95:143-152, 2002

37. Laird JM, Carter AJ, Grauert M, Cervero F: Analgesicctivity of a novel use-dependent sodium channel blocker,robenetine, in mono-arthritic rats. Br J Pharmacol 134:
742-1748, 2001 s
38. Laird JM, Souslova V, Wood JN, Cervero F: Deficits inisceral pain and referred hyperalgesia in Nav1.8 (SNS/PN3)–ull mice. J Neurosci 22:8352-8356, 2002

39. Leffler A, Cummins TR, Dib-Hajj SD, Hormuzdiar WN,lack JA, Waxman SG: GDNF and NGF reverse changes inepriming of TTX-sensitive Na(�) currents following axot-my of dorsal root ganglion neurons. J Neurophysiol 88:650-58, 2002

40. Levine JD, Taiwo YO: Involvement of the mu-opiateeceptor in peripheral analgesia. Neuroscience 32:571-575,989

41. Lindia JA, Kohler MG, Martin WJ, Abbadie C: Relation-hip between sodium channel Nav1.3 expression and neuro-athic pain behavior in rats. Pain 117:145-153, 2005

42. Liu CN, Devor M, Waxman SG, Kocsis JD: Subthresholdscillations induced by spinal nerve injury in dissociateduscle and cutaneous afferents of mouse DRG. J Neuro-hysiol 87:2009-2017, 2002

43. Liu CN, Michaelis M, Amir R, Devor M: Spinal nervenjury enhances subthreshold membrane potential oscilla-ions in DRG neurons: Relation to neuropathic pain. J Neu-ophysiol 84:205-215, 2000

44. Liu CN, Wall PD, Ben-Dor E, Michaelis M, Amir R, Devor: Tactile allodynia in the absence of C-fiber activation:ltered firing properties of DRG neurons following spinalerve injury. Pain 85:503-521, 2000

45. Liu X, Eschenfelder S, Blenk KH, Janig W, Habler H:pontaneous activity of axotomized afferent neurons after5 spinal nerve injury in rats. Pain 84:309-318, 2000

46. Lyu YS, Park SK, Chung K, Chung JM: Low dose of te-rodotoxin reduces neuropathic pain behaviors in an animalodel. Brain Res 871:98-103, 2000

47. MacKinnon R: Voltage sensor meets lipid membrane.cience 306:1304-1305, 2004

48. Malan TP, Ossipov MH, Gardell LR, Ibrahim M, Bian D,ai J, Porreca F: Extraterritorial neuropathic pain correlatesith multisegmental elevation of spinal dynorphin in nerve-

njured rats. Pain 86:185-194, 2000

49. Mao J, Chen L: Systemic lidocaine for neuropathic painelief. Pain 87:7-17, 2000

50. Mao J, Price DD, Mayer DJ, Lu J, Hayes RL: IntrathecalK-801 and local nerve anesthesia synergistically reduce no-

iceptive behaviors in rats with experimental peripheralononeuropathy. Brain Res 576:254-262, 1992

51. Matzner O, Devor M: Hyperexcitability at sites of nervenjury depends on voltage-sensitive Na� channels. J Neuro-hysiol 72:349-359, 1994

52. Maruyama H, Yamamoto M, Matsutomi T, Zheng T,akata Y, Wood JN, Ogata N: Electrophysiological charac-

erization of the tetrodotoxin-resistant Na� channel,a(v)1.9, in mouse dorsal root ganglion neurons. Pflugersrch 449:76-87, 2004

53. Max MB, Lynch SA, Muir J, Shoaf SE, Smoller B, Dubner: Effects of desipramine, amitriptyline, and fluoxetine onain in diabetic neuropathy. N Engl J Med 326:1250-1256,992

54. Michaelis M, Habler HJ, Janig W: Silent afferents: Aeparate class of primary afferents? Clin Exp Pharmacolhysiol 23:99-105, 1996

55. Michiels JJ, Te Morsche RH, Jansen JB, Drenth JP: Auto-
omal dominant erthermalgia associated with a novel mu-

tN

1pa

1Etfl

1pe

1sdP

1StsJ

1tg

1Nsi1

1AD

13p

1Gs

1m1

1Id

1Fp

1tf

1SUspr

1ra1

1

mic

1RLtS9

1m3

1RrtNP

1pp

1t6

1ln

1Ks4

1mcp

1tg

1Ah

1po

1g1

1Lt4

1th

1tta

1C


ation in the voltage-gated sodium channel (alpha) subunitav1.7. Arch Neurol 62:1587-1590, 2005

56. Nassar MA, Levato A, Stirling LC, Wood JN: Neuro-athic pain develops normally in mice lacking both Nav1.7nd Nav1.8. Mol Pain, 2005

57. Nassar MA, Stirling LC, Forlani G, Baker MD, MatthewA, Dickenson AH, Wood JN: Nociceptor-specific gene dele-ion reveals a major role for Nav1.7 (PN1) in acute and in-ammatory pain. PNAS 101:12706-12711, 2004

58. Nietgen GW, Chan CK, Durieux ME: Inhibition of lyso-hosphatidate signaling by lidocaine and bupivacaine. An-sthesiology 86:1112-1119, 1997

59. Nordin M, Nystrom B, Wallin U, Hagbarth K-E: Ectopicensory discharges and paresthesiae in patients with disor-ers of peripheral nerves, dorsal roots and dorsal columns.ain 20:231-245, 1984

60. Novakovic SD, Tzoumaka E, McGivern JG, Haraguchi M,angameswaran L, Gogas KR, Eglen RM, Hunter JC: Distribu-ion of the tetrodotoxin-resistant sodium channel PN3 in ratensory neurons in normal and neuropathic conditions.Neurosci 18:2174-2187, 1998

61. Oda A, Ohashi H, Komori S, Iida H, Dohi S: Characteris-ics of ropivacaine block of Na� channels in rat dorsal rootanglion neurons. Anesth Analg 91:1213-1220, 2000

62. Okuse K, Chaplan SR, McMahon SB, Luo ZD, CalcuttA, Scott BP, Akopian AN, Wood JN: Regulation of expres-

ion of the sensory neuron–specific sodium channel SNS innflammatory and neuropathic pain. Mol Cell Neurosci 10:96-207, 1997

63. Olivera BM, Rivier J, Clark C, Ramilo CA, Corpuz GP,bogadie FC, Mena EE, Woodward SR, Hillyard DR, Cruz LJ:iversity of Conus neuropeptides. Science 249:257-263, 1990

64. Omana-Zapata I, Khabbaz MA, Hunter JC, Bley KR: QX-14 inhibits ectopic nerve activity associated with neuro-athic pain. Brain Res 771:228-237, 1997

65. Orozco OE, Walus L, Sah DW, Pepinsky RB, Sanicola M:FRalpha3 is expressed predominantly in nociceptive sen-

ory neurons. Eur J Neurosci 13:2177-2182, 2001

66. Orstavik K, Mork C, Kvernebo K, Jorum E: Pain in pri-ary erythromelalgia—a neuropathic component? Pain

10:531-538, 2004

67. Pancrazio JJ, Kamatchi GL, Roscoe AK, Lynch C 3rd:nhibition of neuronal Na� channels by antidepressantrugs. J Pharmacol Exp Ther 284:208-214, 1998

68. Pedroarena CM, Pose IE, Yamuy J, Chase MH, MoralesR: Oscillatory membrane potential activity in the soma of arimary afferent neuron. J Neurophysiol 82:1465-1476, 1999

69. Persaud N, Strichartz GR: Micromolar lidocaine selec-ively blocks propagating ectopic impulses at a distancerom their site of origin. Pain 99:333-340, 2002

69b. Pertin M, Ji RR, Berta T, Powell AJ, Karchewski L, TateN, Isom LL, Woolf CJ, Gilliard N, Spahn DR, Decosterd I:pregulation of the voltage-gated sodium channel beta2

ubunit in neuropathic pain models: characterization of ex-ression in injured and non-injured primary sensory neu-ons. J Neurosci 25:10970-10980, 2005

70. Pertovaara A, Wei H, Hamalainen MM: Lidocaine in theostroventromedial medulla and the periaqueductal grayttenuates allodynia in neuropathic rats. Neurosci Lett 218:27-130, 1996

71. Petersen M, Zhang J, Zhang J-M, LaMotte RH. Abnor- p

al spontaneous activity and responses to norepinephrinen dissociated dorsal root ganglion cells after chronic nerveonstriction. Pain 67:391-397, 1996

72. Porreca F, Lai J, Bian D, Wegert S, Ossipov MH, EglenM, Kassotakis L, Novakovic S, Rabert DK, Sangameswaran, Hunter JC: A comparison of the potential role of the te-rodotoxin-insensitive sodium channels, PN3/SNS and NaN/NS2, in rat models of chronic pain. Proc Natl Acad Sci U S A6:7640-7644, 1999

73. Porreca F, Ossipov MH, Gebhart GF: Chronic pain andedullary descending facilitation. Trends Neurosci 25:319-

25, 2002

74. Priest BT, Murphy BA, Lindia JA, Diaz C, Abbadie C,itter AM, Liberator P, Iyer LM, Kash SF, Kohler MG, Kaczo-owski GJ, MacIntyre DE, Martin WJ: Contribution of theetrodotoxin-resistant voltage-gated sodium channelav1.9 to sensory transmission and nociceptive behavior.NAS 102:9382-9387, 2005

75. Puig S, Sorkin LS: Formalin-evoked activity in identifiedrimary afferent fibers: Systemic lidocaine suppresseshase-2 activity. Pain 64:345-355, 1995

76. Puil E, Spigelman I: Electrophysiological responses ofrigeminal root ganglion neurons in vitro. Neuroscience 24:35-46, 1988

76b. Ragsdale DS, McPhee JC, Scheuer T, Catterall WA: Mo-ecular determinants of state-dependent block of Na� chan-els by local anesthetics. Science 265:1724-1728, 1994

77. Raymond CK, Castle J, Garrett-Engele P, Armour CD,an Z, Tsinoremas N, Johnson JM: Expression of alternatively

pliced sodium channel �-subunit genes. J Biol Chem 279:6234-46241, 2004

78. Renganathan M, Cummins TR, Hormuzdiar WN, Wax-an SG: Alpha-SNS produces the slow TTX-resistant sodium

urrent in large cutaneous afferent DRG neurons. J Neuro-hysiol 84:710-718, 2000

79. Renganathan M, Cummins TR, Waxman SG: Contribu-ion of Na(v)1.8 sodium channels to action potential electro-enesis in DRG neurons. J Neurophysiol 86:629-640, 2001

80. Renton T, Yiangou Y, Plumpton C, Tate S, Bountra C,nand P. Sodium channel Nav1.8 immunoreativity in painfuluman dental pulp. BMC Oral Health 5:5, 2005

81. Ritter AM, Mendell LM: Somal membrane properties ofhysiologically identified sensory neurons in the rat: Effectsf nerve growth factor. J Neurophysiol 68:2033-2041, 1992

82. Rowbotham MC, Davies PS, Fields HL: Topical lidocaineel relieves postherpetic neuralgia. Ann Neurol 37:246-253,995

83. Rowbotham MC, Davies PS, Verkempinck C, Galer BS:idocaine patch: Double-blind controlled study of a newreatment method for post-herpetic neuralgia. Pain 65:39-4, 1996

84. Rowbotham MC, Reisner-Keller LA, Fields HL: Both in-ravenous lidocaine and morphine reduce the pain of post-erpetic neuralgia. Neurology 41:1024-1028, 1991

85. Roza C, Laird JMA, Souslova V, Wood JN, Cervero F: Theetrodotoxin-resistant Na� channel Nav1.8 is essential forhe expression of spontaneous activity in damaged sensoryxons of mice. J Physiol 550:921-926, 2003

86. Rugiero F, Mistry M., Sage D, Black JA, Waxman SG,rest M, Clerc N, Delmas P, Gola M: Selective expression of a
ersistent tetrodotoxin-resistant Na� current and NaV1.9

s2

1iv

1Ccr

1tl

1cc1

1Ttrj

1NT

1Sap

1ElP

1mJ

1(L

1ptGa

1rf1

1t5

1Sc1

2lN

2LA

2

c5

2Jdt

2CDuN

2per

2Woar1

2lno2

2tm

2Fts

2fn

2ni

2Lnrp

2lJ

2De

2rt

2is

2nsJ


ubunit in myenteric sensory neurons. J Neurosci3:2715-2725, 2003

87. Rush AM, Waxman SG: PGE2 increases the tetrodotox-n-resistant Nav 1.9 sodium current in mouse DRG neuronsia G-proteins. Brain Res 1023:345-364; 2004

88. Scholz A, Kuboyama N, Hempelmann G, Vogel W:omplex blockade of TTX-resistant Na� currents by lido-aine and bupivacaine reduced firing frequency in DRG neu-ons. J Neurophysiol 79:1746-1754, 1998

89. Sindrup SH, Jensen TS: Efficacy of pharmacologicalreatments of neuropathic pain: An update and effect re-ated to mechanism of drug action. Pain 83:389-400, 1999

90. Sinnott CJ, Garfield JM, Strichartz GR: Differential effi-acy of intravenous lidocaine in alleviating ipsilateral versusontralateral neuropathic pain in the rat. Pain 80:521-531,999

91. Sleeper AA, Cummins TR, Dib-Hajj SD, Hormuzdiar W,yrrell L, Waxman SG, Black JA: Changes in expression ofwo tetrodotoxin-resistant sodium channels and their cur-ents in dorsal root ganglion neurons after sciatic nerve in-ury but not rhizotomy. J Neurosci 20:7279-7289, 2000

92. Sontheimer H, Black JA, Waxman SG: Voltage-gateda� channels in glia: Properties and possible functions.rends Neurosci 19:325-331, 1996

93. Sorensen HJ, Beeler JJ, Johnson LR, Kleier DJ, LevinsonR, Henry MJ: NaV1.7/Pn1 sodium channel upregulation andccumulation at demyelinated sites in painful human toothulp [abstract]. Soc Neurosci 175.13, 2003

94. Stirling LC, Forlani G, Baker MD, Wood JN, MatthewsA, Dickenson AH, Nassar MA: Nociceptor-specific gene de-etion using heterozygous Nav1.8–Cre recombinase mice.ain 113:27-36, 2005

94b. Strichartz GR: The inhibition of sodium currents inyelinated nerve by quaternary derivatives of lidocaine.Gen Physiol 62:37-57, 1973

95. Strichartz GR, Berde CB: Local anesthetics, in Miller RDed): Miller’s Anesthesia. Philadelphia, PA, Elsevier Churchillivingstone, 2005, pp 573-603

96. Strichartz GR, Zhou Z, Sinnott C, Khodorova A: Thera-eutic concentrations of local anaesthetics unveil the poten-ial role of sodium channels in neuropathic pain, in Bock G,oode JA (eds); Sodium Channels and Neuronal Hyperexcit-bility, Hoboken, NJ, John Wiley & Sons, 2002, pp 189-205

97. Su X, Sengupta JN, Gebhart GF: Effects of kappa opioideceptor–selective agonists on responses of pelvic nerve af-erents to noxious colorectal distension. J Neurophysiol 78:003-1012, 1997

98. Sudoh Y, Cahoon EE, Gerner P, Wang GK: Tricyclic an-idepressants as long-lasting local anesthetics. Pain 103:49-5, 2003

99. Sullivan LM, Honemann CW, Arledge JA, Durieux ME:ynergistic inhibition of lysophosphatidic acid signaling byharged and uncharged local anesthetics. Anesth Analg 88:117-1124, 1999

00. Sunami A, Dudley SC, Fozzard HA: Sodium channel se-ectivity filter regulates antiarrhythmic drug binding. Procatl Acad Sci U S A 94:14126-14131, 1997

01. Takao Y, Mikawa K, Nishina K, Maekawa N, Obara H:idocaine attenuates hyperoxic lung injury in rabbits. Actanaesthesiol Scand 40:318-325, 1996

02. Tal M, Devor M: Ectopic discharge in injured nerves: 2

omparison of trigeminal and somatic afferents. Brain Res79:148-151, 1992

03. Tanaka M, Cummins TR, Ishikawa K, Dib-Hajj SD, BlackA, Waxman SG: SNS Na� channel expression increases inorsal root ganglion neurons in the carrageenan inflamma-ory pain model. Neuroreport 9:967-972, 1998

04. Tate S, Benn S, Hick C, Trezise D, John V, Mannion RJ,ostigan M, Plumpton C, Grose D, Gladwell Z, Kendall G,ale K, Bountra C, Woolf CJ: Two sodium channels contrib-te to the TTX-R sodium current in primary sensory neurons.at Neurosci 1:653-655, 1998

05. Toledo AJ, Brehm P, Halegoua S, Mandel G: A singleulse of nerve growth factor triggers long-term neuronalxcitability through sodium channel gene induction. Neu-on 14:607-611, 1995

06. Toledo-Aral JJ, Moss BL, He ZJ, Koszowski AG,hisenand T, Levinson SR, Wolf JJ, Silos-Santiago I, Haleg-

ua S, Mandel G: Identification of PN1, a predominant volt-ge-dependent sodium channel expressed principally in pe-ipheral neurons. Proc Natl Acad Sci U S A, 94:1527-1532,997

07. Tu H, Deng L, Sun Q, Yao L, Han JS, Wan Y: Hyperpo-arization-activated, cyclic nucleotide–gated cation chan-els: Roles in the differential electrophysiological propertiesf rat primary afferent neurons. J Neurosci Res 76:713-722,004

08. Vandam LD: Some aspects of the history of local anes-hetics, in Strichartz GR (ed): Local Anesthetics, Berlin, Ger-any, Springer-Verlag, 1987, pp 1-12

09. Villarreal CF, Sachs D, de Queiroz Cunha F, Parada CA,erreira SH: The role of Nav1.8 sodium channel in the main-enance of chronic inflammatory hypernociception. Neuro-ci Lett 2005;386:72-77

10. Wall PD, Devor M: Sensory afferent impulses originaterom dorsal root ganglia as well as from the periphery inormal and nerve-injured rats. Pain 17:321-339, 1983

11. Wall PD, Gutnick M: Ongoing activity in peripheralerves: The physiology and pharmacology of impulses orig-

nating from a neuroma. Exp Neurol 43:580-593, 1974

12. Wang R, Guo W, Ossipov MH, Vanderah TW, Porreca F,ai J. Glial cell line–derived neurotrophic factor normalizeseurochemical changes in injured dorsal root ganglion neu-ons and prevents the expression of experimental neuro-athic pain. Neurosci 121:815-24, 2003

13. Wang DW, Nie L, George AL Jr, Bennett PB: Distinctocal anesthetic affinities in Na� channel subtypes. Biophys70:1700-1708, 1996

14. Wang SY, Nau C, Wang GK: Residues in Na(�) channel3-S6 segment modulate both batrachotoxin and local an-sthetic affinities. Biophys J 79:1379-1387, 2000

15. Wang SY, Wang GK: Point mutations in segment I-S6ender voltage-gated Na� channels resistant to batracho-oxin. Proc Natl Acad Sci U S A 95:2653-2658, 1998

16. Watkins LR, Maier SF: Beyond neurons: Evidence thatmmune and glial cells contribute to pathological paintates. Physiol Rev 82:981-1011, 2002

17. Waxman SG, Kocsis JD, Black JA: Type III sodium chan-el mRNA is expressed in embryonic but not adult spinalensory neurons, and is reexpressed following axotomy.Neurophysiol 72:466-470, 1994

18. Wood JN, Boorman JP, Okuse K, Baker MD: Voltage-

g6

2i

2gt

2ia7

2TDn7

2an3

2to1

2at1

2bng

2tR

2i

C5

2td

2mdr3

2BSt2

2svN2

2ea3

2sTr

2Skig2

2lin


ated sodium channels and pain pathways. J Neurobiol1:55-71, 2004

19. Woolf CJ: Evidence for a central component of post-njury pain hypersensitivity. Nature 306:686-688, 1983

20. Woolf CJ, Safieh GB, Ma QP, Crilly P, Winter J: Nerverowth factor contributes to the generation of inflamma-ory sensory hypersensitivity. Neuroscience 62:327-331, 1994

21. Wright SN, Wang SY, Kallen RG, Wang GK: Differencesn steady-state inactivation between Na channel isoformsffect local anesthetic binding affinity. Biophys J 73:779-88, 1997

22. Wu G, Ringkamp M, Murinson BB, Pogatzki EM, HartkeV, Weerahandi HM, Campbell JN, Griffin JW, Meyer RA:egeneration of myelinated efferent fibers induces sponta-eous activity in uninjured C-fiber afferents. J Neurosci 22:746-7753, 2002

23. Wu N, Hsiao CF, Chandler SH. Membrane resonancend subthreshold membrane oscillations in mesencephalic Veurons: Participants in burst generation. J Neurosci 21:729-3739, 2001

24. Xiao WH, Bennett GJ: Gabapentin has an antinocicep-ive effect mediated via a spinal site of action in a rat modelf painful peripheral neuropathy. Analgesia 2:267-273,996

25. Xiao WH, Yu AL, Sorkin LS: Electrophysiological char-cteristics of primary afferent fibers after systemic adminis-ration of antiGD2 ganglioside antibody. Pain 69:145-151,997

26. Xing JL, Hu SJ, Jian Z, Duan JH: Subthreshold mem-rane potential oscillation mediates the excitatory effect oforepinephrine in chronically compressed dorsal root gan-lion neurons in the rat. Pain 105:177-183, 2003

27. Xing JL, Hu SJ, Long KP: Subthreshold membrane po-ential oscillations of type A neurons in injured DRG. Braines 901:128-136, 2001

28. Xiong Z, Bukusoglu C, Strichartz GR: Local anesthetics
nhibit the G protein-mediated modulation of K� and J
a�� currents in anterior pituitary cells. Mol Pharmacol5:150-158, 1999

29. Yang D, Gereau RW 4th: Peripheral group II metabo-ropic glutamate receptors mediate endogenous antiallo-ynia in inflammation. Pain 106:411-417, 2003

30. Yang D, Gereau RW 4th: Group II metabotropic gluta-ate receptors inhibit cAMP-dependent protein kinase–me-iated enhancement of tetrodotoxin-resistant sodium cur-ents in mouse dorsal root ganglion neurons. Neurosci Lett57:159-162, 2004

31. Yang Y, Wang Y, Li S, Xu Z, Li H, Ma L, Fan J, Bu D, Liu, Fan Z, Wu G, Jin J, Ding B, Zhu X, Shen Y: Mutations inCN9A, encoding a sodium channel alpha subunit, in pa-ients with primary erythermalgia. J Med Genet 41:171-174,004

32. Yoshimura N, Seki S, Novakovic SD, Tzoumaka E, Erick-on VL, Erickson KA, Chancellor MB, de Groat WC: The in-olvement of the tetrodotoxin-resistant sodium channela(v)1.8 (PN3/SNS) in a rat model of visceral pain. J Neurosci1:8690-8696, 2001

33. Zhang YH, Nicol GD: NGF-mediated sensitization of thexcitability of rat sensory neurons is prevented by a blockingntibody to the p75 neurotrophin receptor. Neurosci Lett66:187-192, 2004

34. Zhang YH, Vasko MR, Nicol GD: Ceramide, a putativeecond messenger for nerve growth factor, modulates theTX-resistant Na(�) current and delayed rectifier K(�) cur-ent in rat sensory neurons. J Physiol 544:385-402, 2002

35. Zhang XF, Zhu CZ, Thimmapaya R, Choi WS, Honore P,cott VE, Kroeger PE, Sullivan JP, Faltynek CR, Gopala-rishnan M, Shieh CC: Differential action potentials and fir-ng patterns in injured and uninjured small dorsal root gan-lion neurons after nerve injury. Brain Res 1009:147-158,004

36. Zhou Z, Davar G, Strichartz G: Endothelin-1 (ET-1) se-ectively enhances the activation gating of slowly inactivat-ng tetrodotoxin-resistant sodium currents in rat sensoryeurons: A mechanism for the pain-inducing actions of ET-1.
Neurosci 22:6325-6330, 2002

The Role of Sodium Channels in Chronic …...in voltage-gated sodium channels may play a role in...

Documents

Transcript of The Role of Sodium Channels in Chronic …...in voltage-gated sodium channels may play a role in...